But before you grab your axe and scramble for the nearest beanstalk, it’s important to survey the scene and make sure that you’re heading for a cash cow rather than being left holding a handful of dry beans.

Metaphors duly stretched, today we take a closer look at the opportunities this presents for the channel with Gilles Closset, our global AI ecosystem leader. Gilles is instrumental in helping our partner community distinguish between the hype and the practical opportunities AI offers and brings a wealth of customer insights.

Gilles, thanks for joining us – let’s kick off.

Q1. Which industries are using AI the most: who are the early adopters? Why are they using it and what business outcomes are they looking for?  

GC: Healthcare is one sector where we’re seeing huge growth, largely because of AI’s potential to help people live longer, take better care of them and reduce administrative burdens. A lot of AI in healthcare doesn’t depend on GenAI, and we see a lot of startups in this area, mostly based more on the traditional machine learning industry.

We also see significant AI uptake in Manufacturing, mostly in the optimisation of production processes. With its heritage in Kaizen and Six Sigma, manufacturing has always seen a strong appetite in this area. The sector is also starting to embrace predictive maintenance, using both digital twin and machine learning technology.

Thirdly,the Retail industry has embraced a mix of GenAI and machine learning to improve customer interactions using chatbots, sales optimisation and hyper-personalisation. AI is allowing every brand to offer the kind of personalised experience that was previously restricted to luxury brands; it has the power to absorb a lot of data points about customers and provide a far more customised experience across the areas of both sales and customer support.

Finally, we see a great amount of AI adoption in the Financial sector, particularly in the areas of fraud detection and personalised engagement. The financial sector has historically been enthusiastic about machine learning for improving KYC (Know Your Customer) processes, and AI can really accelerate this.

Q2. What are other customers really looking for when it comes to AI? What are they adopting, and what are the best opportunities for VARs and MSPs in AI today?

GC: We see a lot of potential from AI to improve efficiency and personalisation, draw together data from different customer journeys and integrate it. As I said above, customers are increasingly expecting a highly integrated approach when it comes to their relationship with brands – and AI enables everyone to do this.

We also hear a lot of questions from customers about data handling; from ethical principles to cloud audits and hands-on management, there’s a lot of opportunity for partners to support customers. It’s still early days in terms of adoption, but this is a clear opportunity for partners to get ahead, show differentiation and stand out.

Q3. Turning to the market main challenges, what themes are customers struggling with the most and need help from VARs and MSPs to support them?

GC: One major challenge we’ve seen is to go beyond a POC – we see a lot of exploratory initiatives, but then they don’t scale!

At the same time, data quality is a challenge. A lot of organisations have legacy systems that weren’t designed to work with AI, so there’s a real need for modernisation. IDC’s latest findings predict an unprecedented YoY growth of 97% in AI compute and storage hardware infrastructure to support this adoption. This area already is a vast area of investment, currently valued at $47.4b. We see a lot of projects to update and extract data from old systems, and partners who can support with this are in a position to significantly drive their customers forward in terms of AI maturity.

Finally – and this is linked to the previous point: we are seeing a big talent gap. AI is a new industry, so organisations are still developing their skills. There are two main ways to address this: you can either help a customer to address the gap directly and deliver workshops and training, or you can provide the end customer with people to do the work on a temporary basis.

All of this requires a transformation plan and can be daunting, but it does represent a major opportunity to add value. Furthermore, these projects will usually start with data literacy, making sure people understand the potential of AI, getting buy-in, and then you’re off!

Q4. What are your top tips for partners starting to explore the possibilities of AI for their customers?

GC:

1: Mind the skills gap: Upskill your own team or help to skill up the customer.

2: Know how to tackle topics around data quality and governance. Good data is the foundation for all good AI – and as the old saying goes, Garbage In, Garbage Out.

3: Use UK or European organisations where you can. We see lots of feedback from partners who are trying to differentiate from the hyperscalers in terms of price, data protection (especially around sovereignty and vulnerability to extra-territorial laws) and working with a European player will make this much easier.

Gilles, many thanks for your time, and very useful insights. From what we’ve heard, AI certainly offers prospects of gargantuan proportions and, like many new rapid-growth market cycles, once you’ve separated the wheat from the chaff (or beans from the stalks), climbing up to the giant’s AI tech lair is a significant opportunity for partners. As always, it’s important to listen to what customers are doing and understand their broader strategic and technical landscape, but in the long term, embracing the AI opportunity can uncover new business treasures you never knew existed!

AI Act regulation and actual implementation at a larger scale is only a few weeks ahead with its first real impact on businesses.

We wanted to recap the main information to have in mind with this short recap – as of April 2025.

𝗘𝗨 𝗔𝗜 𝗔𝗰𝘁: 𝘀𝗲𝗰𝘂𝗿𝗶𝘁𝘆 𝗳𝗮𝗰𝘁𝗼𝗿 𝗼𝗿 𝗰𝗼𝗻𝘀𝘁𝗿𝗮𝗶𝗻𝘁 𝗳𝗼𝗿 𝘁𝗵𝗲 𝗔𝗜 𝗺𝗮𝗿𝗸𝗲𝘁 𝗶𝗻 𝗘𝘂𝗿𝗼𝗽𝗲?

The EU’s AI Act is a global regulation already implemented in Europe, with now the Ban on AI Systems with Unacceptable risks since February 2025.

The AI Act applies to providers and deployers of AI systems within the EU, aiming to ensure product safety and mitigate risks associated with AI use.

🔑 Core objectives:
• Creating a framework of trust for AI adoption
• Encouraging responsible innovation
• Protecting fundamental rights and user data

⏱️ Key timeline milestones:
• August 2024: Law enters into force
• February 2025: Ban on unacceptable risk AI systems
• May 2025: AI Office Codes of Practice ready
• August 2025: GPAI model regulations effective
• August 2026: Full application to all AI systems

The Act creates a two-level governance structure:
1️ National authorities supervising AI systems
2️ European Commission and AI Office regulating general-purpose AI models

While the legal complexity is significant, the goal is straightforward: making AI trustworthy while fostering innovation.

Businesses face several challenges in complying with the EU AI Act due to its complexity and stringent requirements.

𝗪𝗵𝗶𝗰𝗵 𝗹𝗲𝘃𝗲𝗹 𝗼𝗳 𝗿𝗶𝘀𝗸 𝗰𝗼𝗻𝗰𝗲𝗿𝗻𝘀 𝘆𝗼𝘂𝗿 𝗔𝗜 𝘀𝘆𝘀𝘁𝗲𝗺 ?

The EU AI Act introduces a four-tier risk classification system that is key to understand:

🔴 Unacceptable Risk (PROHIBITED): • Social scoring systems • Emotion recognition in workplaces/education • Biometric categorization systems that infer sensitive attributes • Scraping facial images from the internet • Manipulative or deceptive AI
🟠 High Risk: • Medical devices • Recruitment systems • Critical infrastructure • Education and vocational training • Law enforcement systems
🟡 Specific Transparency Risk: • Chatbots • Deepfakes • AI-generated content → Must clearly disclose they are AI-generated
🟢 Minimal Risk: • Anti-spam filters • AI in video games • Most business applications

The penalties are severe: up to 7% of global annual turnover or €35 million for violations related to prohibited practices.

Before deploying any AI system, you need to assure the use of the AI system and to assess its risk – this will determine your compliance exigence.

Talking about compliance is also talking about processes and governance.
OVHcloud and ETHIQAIS have been among the very first signatories of the AI Pact at the European Commission and can help you navigate in this new regulatory framework:

With OVHcloud:
🔹 European Trusted Cloud with Data Sovereignty at its heart
🔹 A wide rage of GPU & Compute capabilities with best-in-class sustainable infrastructure

With Ethiqais:
🔹 Automate documentation for AI systems
🔹 Elaborate AI global governance and real-time conformity assessment

Know Your Role & Responsibilities

Your obligations under the EU AI Act depend on your role in the AI value chain. The 2 main categories are:

👨‍💻 𝗦𝘂𝗽𝗽𝗹𝗶𝗲𝗿/𝗗𝗲𝘃𝗲𝗹𝗼𝗽𝗲𝗿:

• You develop AI systems and put them on the market

• Highest level of obligations, especially for high-risk systems

• Responsible for risk assessments, documentation, technical requirements

👩‍💼 𝗗𝗲𝗽𝗹𝗼𝘆𝗲𝗿/𝗨𝘀𝗲𝗿:

• You use AI systems in professional contexts

• Must follow instructions from suppliers

• Additional obligations if using high-risk systems

🤝 Other roles with specific obligations: • Agent: EU representative for non-EU suppliers • Importer: Brings AI systems into the EU market • Distributor: Makes AI systems available on the EU market

See the attached table for an overview of obligations status per role

And check “The AI Act Explorer” provided by the Future of Life Institute for all the related details.

Start preparing NOW by:

1 – Identifying all AI systems in your organization

2 – Determining your role for each system

3 – Classifying each system’s risk level

4 – Understanding your specific obligations

5 – Implementing compliance measures before deadlines

And now… Time to introduce more formally ETHIQAIS, a member of the OVHcloud Startup Program!

What tool for AiAct in France and Europe? ETHIQAIS – your platform for the global AI management, governance and compliance of artificial intelligence systems

Why do you need to define the governance of artificial intelligence systems within your company?

Like any technology, artificial intelligence needs to be challenged in the way it is implemented, in order to evolve in a way that is consistent with human needs. AI addresses issues linked to climate change, energy saving and production, education, health, transport, finance and many other sectors.

Adopting AI in an existing product, service or system, or in the ideation phase, requires defining a structure, a process, and identifying the resources and tools needed throughout the AI lifecycle.

Managing and monitoring the AI system has become essential in order to control risks and ensure that AI is aligned with current regulations.

All players involved in the AI lifecycle are required to participate in the development of its governance: managers, but also technical teams – AI developers, data scientists, product owners, product managers, CIOs, CDOs, quality managers, AI lawyers, compliance officers, etc. Governance doesn’t just refer to company policy, it’s the first step in operationalizing AI in company: defining the rules for adopting AI internally, standardizing and structuring documentation production to ensure transparency and collaborative decision-making, adopting AI management and monitoring tools to foster a faster ROI and to limit reputational, legal but also economic and technical risks.”

This approach requires the adoption of a global AI management solution. ETHIQAIS and OVHcloud are working together to offer companies concrete added value in the development of AI governance, from the ideation phase right through to market launch. As guarantors of security, compliance, robustness and technological sovereignty, ETHIQAIS and OVHcloud are joining forces to provide companies developing AI with the ETHIQAIS platform.

ETHIQAIS is an intelligent platform designed to help developers of AI systems and models, data scientists to create automated AI documentation, and managers to define AI governance while complying with regulations, notably the European AI Act.

ETHIQAIS automatically collects data to manage and document your organization’s AI systems. ETHIQAIS continuously monitors your systems in an automated way to ensure you remain compliant with the European AI Act Regulation. The traceability of your AI systems and the monitoring of your use cases is an essential element for compliance, but above all for steering your product roadmap and managing your organization.

In addition to ensuring the traceability of AI systems right from their production, ETHIQAIS develops functionalities for analyzing the requirements and controls set out in standards and regulations. ETHIQAIS is designed to detect irregularities that might escape the analysis of your teams, and to suggest improvements to reduce risks.

Do not take any risks, request a demo quickly to help you build your skills and master the development of your AI system.

– A demo: contact@ethiqais.com
– More about ETHIQAIS: ethiqais.com

ETHIQAIS and OVHcloud provide you with the essential tools to help you build trust and scale your AI products in a responsible, secure and compliant way.

New and innovative capabilities come from all parts of the technology ecosystem and from all regions of the world. Every week, almost every day!

The momentum in this space is incredible. In fact, we’ve seen a significant acceleration in the number of AI startups that have joined the OVHcloud Startup Program as well as Partners & Editors adding AI expertise & capabilities to their portfolio.

Aligned with our DNA of a Trusted & Sustainable Cloud, OVHcloud is committed to supporting AI innovation that adheres to core values.

To help customers and developers harness this innovation, we’re bringing the best of OVHcloud’s infrastructure, AI products, and State-of-the-art models to members of our Ecosystem at every layer of the AI stack: chipmakers, models builders and AI platforms, technology partners enabling companies to develop and deploy machine learning (ML) models, app-builders solving customer use-cases with generative AI, and global services and consulting firms that help enterprise customers implement all of this technology at scale.

Let’s deep dive into our partnerships, programs, and resources for each segment of the ecosystem that showcase our open approach.

Building a Trusted and Innovative AI Ecosystem

Model builders & Chipmakers

Let’s kickstart the introduction of our AI Ecosystem with the members that are directly integrated into specific OVHcloud AI products, aka Technology Partners.

Companies like Mistral AI, Meta and Stability AI are building open-source foundation models, including LLMs, that can significantly accelerate the development of generative AI and natural language processing (NLP) applications. OVHcloud serves to end-customers these models through AI Endpoints with its high-performance infrastructure and industry-leading energy efficiency.

AI endpoints require no AI expertise or dedicated infrastructure, as the serverless platform provides access to advanced AI models including Large Language Models (LLMs), natural language processing, translation, speech recognition, image recognition, and more. Developers can select from a range of models, including open-source options like Mistral AI, Llama, Whisper, and Stable Diffusion, as well as a variety of optimized models from our Model Builders partners, creating a versatile testing ground for chosen AI models.

Our catalog of AI models is continually expanding, and we are actively seeking new collaborations with partners to integrate proprietary models that address specific use cases.

OVHcloud also developed strong and long-lasting partnerships with chipmakers like NVIDIA and AMD to deliver tailored services for deep learning, inference and high-performance computing, with the best available GPUs. AI models are becoming more complex due to the rise of conversational AI. Training and inference now require massive computing power and scalability, and OVHcloud follows the industry innovations by integrating the latest GPUs, including for 2025 the AMD MI325X series, and the Nvidia H200 NVL and Blackwell generation. Using industrial innovations, such as water cooling in our servers, allow us to achieve the lowest energy consumption on the market.

AI PaaS Solutions & Tools

Organizations and developers engaged in ambitious AI projects usually employ various tools to facilitate the creation, management, and deployment of their models. These tools assist developers with essential tasks such automating and optimizing data pipelines, monitoring model performance, managing private datasets, defining and enforcing safety & security measures related to regulation or specific policies. OVHcloud collaborates with these organizations to address the crucial requirements of machine learning engineers and data scientists.

To meet growing demand from customers and partners building innovative AI services on OVHcloud, many of the leaders in AI solutions are launching new or expanded partnerships with OVHcloud today. Let’s have a look to these few examples:

• Multiverse Computing are the world leaders in Quantum AI. They apply quantum and quantum-inspired AI to solve complex problems delivering practical applications and tangible value today.
  
• Hopsworks seamlessly integrates and can be deployed on OVHcloud using Kubernetes, allowing users to run feature engineering pipelines, training pipelines, and batch inference pipelines using Spark, Flink, or Python on OVHcloud.
  
• With Valohai access scalable and secure cloud environments without having to rebuild your ML workflows. The integration between the Valohai MLOps platform and OVHcloud makes it easy to access on-demand computational resources. Scale up with ease to meet the needs of your projects, while ensuring data security and regulatory compliance
  
• Lampi AI provides a Secure AI platform with the best and latest LLMs to power predictable and fine-tuned AI agents that pick the relevant information from your data and web, reason, iterate, and tackle complex tasks.
  
• Qdrant : Through the seamless integration between Qdrant Hybrid Cloud and OVHcloud, developers and businesses are able to deploy the fully managed vector database within their existing OVHcloud setups in minutes, enabling faster, more accurate AI-driven insights.

Through our support to these critical members, we offer developers the best platform and ecosystem in which to build the next generation of helpful AI applications, and provide customers with a single destination for building, innovating with, and applying AI.

AI Apps addressing End-customers use cases specifically

OVHcloud is the destination for developers and partners to build the next generation of innovative applications with AI and ML, including exciting new generative AI capabilities.

Much innovation in the generative AI space comes from fast moving, early-stage startups. They excel in developing new applications designed to address very specific End-customers’ use cases. Some differentiate through their model(s), either proprietary or fined-tuned, and make it available through inference API or in their App. Others bring value buy developing Applications or User Interface on top of “General-Purpose AI Models” – so-called API wrappers – by knowing precisely the business workflow of their customers. This may translate into AI agents capable of performing tasks independently, without the need for constant human oversight.

Many AI startups are choosing OVHcloud not only for industry-leading Sustainable & Trusted Cloud infrastructure, but also for fully managed AI services, which makes it faster and easier to scale, at the best price and with no lockin.

Let’s review some of them:

• ILLUIN Technology provides a powerful low-code multimodal AI orchestration platform that enables you to hybridize different AI approaches and models to implement and industrialize your most complex customized use cases, including AI Agents.
  
• moinAI uses AI to automatically resolve recurring customer enquiries – across multiple channels and in various languages – with minimal effort. Chatbots, live chat and product advisors allow companies to communicate quickly and efficiently with customers on the website around the clock.
  
• catchHR uses AI to streamline recruitment, automating tasks like job posting, candidate sourcing, and skill matching to save time and boost efficiency. AI-generated job ads attract top talent, while AI-powered candidate analysis ensures a strong match between applicants and roles, considering both skills and personality fit.
  
• Rayscape has already demonstrated excellent results in 150+ clinics and hospitals. Its AI is trained on more than 43 million images from all around the globe and powers predictive insights, automated analysis, efficient workflows to prioritize cases based on urgency, and generates structured reports.
  
• Factiverse offers AI-powered solutions to enhance content credibility and streamline fact-checking processes. Their offerings include an advanced text editor that identifies and verifies factual statements in your content. It highlights claims and provides links to credible sources, assisting in correcting inaccuracies Factiverse GPT.

Services Partners

We stand at the brink of an exhilarating transformation, propelled by advancements in machine learning (ML) technologies. This shift holds the promise of revolutionizing customer experiences, introducing groundbreaking applications, and boosting our customers’ productivity to new levels. The market’s enthusiasm is clear, with an unprecedented number of customers eager to leverage generative artificial intelligence to revamp their businesses.

Successfully innovating with large language models and generative AI demands proficiency in data management, AI, human resources, and operational processes. It is crucial that these models and AI solutions are crafted to be ethical, transparent, and reliable.

Our partner ecosystem will lead the way in developing innovative business solutions tailored to customers across various industries and sizes.

Services partners from our Ecosystem have demonstrated expertise delivering Machine Learning and generative AI solutions on OVHcloud. These partners offer a range of products and services and technologies including specialized consulting services, Managed Services and Applications that are secure, efficient, and scalable across industries.

Today, several of our leading partners, CGI, Onepoint, Accenture, synaigy , W&B Asset Studio, Sopra Steria, Inetum and NEXiD are already providing key support in terms of OVHcloud generative AI advisory, implementation services and capabilities available to customers. These partners play an essential role in applying new AI capabilities to solve industry-specific challenges and helping enterprises build generative AI into their products and everyday business processes.

OVHcloud and its broader ecosystem

Our AI Ecosystem is part of our broader Ecosystem and includes a wide variety of startups and partners, ready to support customers with both current and future technological challenges. It does so by giving customers the means to innovate and develop their own competitive advantage.

Through these programs, we provide product support, marketing amplification, and co-selling opportunities to help our services and ISV partners bring these solutions to market faster, reach more customers, and grow their businesses.

We have launched over the past 10+ years the following initiatives:

OVHcloud Partner Program

OVHcloud partners play a key role in customers’ digital transformation, with the support and services they offer to help them meet the challenges involved. Over 700 companies joint this program, providing a wide range of expertise and services to our customers.

Interested Partners can go here to apply to join the program

OVHcloud Startup Program

We nurture tech entrepreneurs by deploying an array of business scaling opportunities within OVHcloud’s global ecosystem of trust.

Over 5.000 startups and scaleups from across the globe that have already benefited from our program since its launch in 2015.

To further support the AI startups and accelerate their app development, we’re launching a new initiative, called AI Accelerator which recognizes select startups whose applications and platforms are optimized to run as-a-service on OVHcloud infrastructure and who are utilizing OVHcloud’s AI capabilities in new and helpful ways. The program provides dedicated access to OVHcloud expertise, training, and co-marketing support to help partners build capacity and go to market.

Interested AI startups can go here to apply to join the program

Open Trusted Cloud  

This program is aimed at software publishers, as well as SaaS and PaaS solution providers. Its ambition is to work together on building an ecosystem of SaaS and PaaS services — hosted in the open, reversible and trusted cloud offered by OVHcloud. This will provide a common platform for competitive solutions, and hundreds have already joined.

You can browse some of the solutions available in our ecosystem here

OVHcloud Marketplace

At the heart of the ecosystem, the Marketplace was designed to benefit everyone. OVHcloud Marketplace brings together the best solutions from SaaS and PaaS publishers in the ecosystem on an ethical and transparent cloud. Carry out the digital transformation of your company or subscribe to a solution for your personal use with complete peace of mind thanks to these trusted solutions.

Technology partners

The OVHcloud vision is to create a transparent, reversible and interoperable cloud. We work with the best players on the market to deliver solutions for the most high-performance, high-security requirements.

OVHcloud is committed to democratize AI within organizations, through a wide range of solutions positioned at every price point while advocating for Digital Sovereignty & Sustainability.

Should your company consider to leverage OVHcloud, would like to know more about our vibrant AI Ecosystem or share a feedback, please feel free to contact me!

Choosing the Right Method for External Knowledge

In AI development, incorporating proprietary data and external knowledge is crucial. Two key methodologies are Retrieval Augmented Generation (RAG) and fine-tuning. Here’s a quick comparison.

𝐑𝐞𝐭𝐫𝐢𝐞𝐯𝐚𝐥 𝐀𝐮𝐠𝐦𝐞𝐧𝐭𝐞𝐝 𝐆𝐞𝐧𝐞𝐫𝐚𝐭𝐢𝐨𝐧 (𝐑𝐀𝐆) 🔍

RAG combines an LLM’s reasoning with external knowledge through three steps:
1️⃣ Retrieve: Identify related documents from an external knowledge base.
2️⃣ Augment: Enhance the input prompt with these documents.
3️⃣ Generate: Produce the final output using the augmented prompt.

The retrieve step is pivotal, especially when dealing with large knowledge bases. Vector databases are often used to manage and search these extensive datasets efficiently.
Implementing a RAG-Chain with Vector Databases: time to recall the post “AI concept in a Nutshell: LLM series – Embeddings & Vectors ” from 1 month ago!

𝐅𝐢𝐧𝐞-𝐓𝐮𝐧𝐢𝐧𝐠 🛠️

Fine-tuning adjusts the LLM’s weights using proprietary data, extending its capabilities to specific tasks.

Approaches to Fine-Tuning:
1️⃣ Supervised Fine-Tuning: Uses demonstration data with input-output pairs.
2️⃣ Reinforcement Learning from Human Feedback: Requires human-labeled data and optimizes the model based on quality scores.

Both approaches need careful decision-making and can be complex.

𝐖𝐡𝐞𝐧 𝐭𝐨 𝐔𝐬𝐞 𝐑𝐀𝐆 𝐨𝐫 𝐅𝐢𝐧𝐞-𝐓𝐮𝐧𝐢𝐧𝐠? 🤔

· RAG: Great for adding factual knowledge without altering the LLM. Easy to implement but adds extra components.
· Fine-Tuning: Best for specializing in new domains. Offers full customizability but requires labeled data and expertise. May cause catastrophic forgetting.

Choose based on your needs and resources. Both methods have their strengths and challenges, making them valuable tools in AI development.

LLM Temperature

I love the analogy shared during one of the break-out sessions at the OVHcloud summit: LLM temperature is like blood alcohol level – the higher it is, the more unexpected the answers! 😂

To be more specific, temperature is a parameter that controls the randomness of the model’s output. A higher temperature encourages more diverse and creative responses, while a lower temperature makes the output more deterministic and predictable.

🔍 Key Points:
🔹 High Temperature: More random and creative outputs, useful for brainstorming and generating novel ideas.
🔹 Low Temperature: More predictable and coherent outputs, ideal for factual information and structured content.

Understanding and adjusting the temperature can help tailor LLM outputs to specific needs, whether you’re looking for creativity or precision.

Low-rank adaptation (LoRA): Turning LLMs and other foundation models into specialists

Before a foundation model is ready to take on real-world problems, it’s typically fine-tuned on specialized data and its billions, or trillions, of weights are recalculated. This style of conventional fine-tuning is slow & expensive.

⚡ LoRA is a quicker solution. With LoRA, you fine-tune a small subset of the base model’s weights, creating a plug-in module that gives the model expertise in, for example, biology or mathematical reasoning at inference time. Like custom bits for a multi-head screwdriver, LoRAs can be swapped in and out of the base model to give it specialized capabilities.

By detaching model updates from the model itself, LoRA has become the most popular of the parameter-efficient fine-tuning (PEFT) methods to emerge with generative AI.

The LoRA approach also makes it easier to add new skills and knowledge without overwriting what the model previously learned, a phenomenon known as catastrophic forgetting. LoRA offers a way to inject new information into a model without sacrificing performance.

But perhaps its most powerful benefit comes at inference time. Loading LoRA updates on and off a base model with the help of additional optimization techniques can be much faster than switching out fully tuned models. With LoRA, hundreds of customized models or more can be served to customers in the time it would take to serve one fully fine-tuned model.

Chain of Thought (CoT)

One of the most effective techniques to improve the performance of a prompt is the Chain of Thought (CoT).

The principle consists of breaking down a problem into several tasks or summarizing all the data available upstream.

Examples:

1) 𝗖𝗼𝗻𝘁𝗲𝗻𝘁 𝗴𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝗼𝗻: Rather than directly asking the LLM to generate a summary, we can first ask it to generate a table of contents and then a summary.

The final result will therefore have a better chance of being exhaustive on the content of the initial document.

2) 𝗖𝗿𝗲𝗮𝘁𝗶𝗼𝗻 𝗼𝗳 𝘀𝘁𝗿𝘂𝗰𝘁𝘂𝗿𝗲𝗱 𝗲𝗻𝘁𝗶𝘁𝗶𝗲𝘀: The CoT allows you to control the format of generation of structured entities (for example in a YAML format that we will have detailed).

These examples will greatly reduce the chances of having hallucinations in the final response of the LLM.

3) 𝗔𝗴𝗲𝗻𝘁 𝗪𝗼𝗿𝗸𝗳𝗹𝗼𝘄: For an LLM Agent, it is preferable to ask the LLM to explicitly generate a plan of the actions it must perform to accomplish its task.

“Start by making a plan of the actions you will have to perform to solve your task. Then answer with the tools you want to use in a YAML block, example: […]”

4) 𝗥𝗔𝗚 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝗼𝗻: Using CoT can improve the quality of the generated response while reducing hallucinations. Asking the LLM to summarize the relevant information contained in the documents will strengthen this semantic field when it has to generate the final response.

Conclusion

The chain of thought allows to strengthen the semantic field of the expected result in two different ways:

• Explicit chain of thought: by giving examples of the expected result in the prompt

• Internal chain of thought: by telling the LLM to generate intermediate “reasoning” steps

This increases latency and costs but in many situations the result is worth the investment

LLM Quantization

Quantization in AI is a technique used to reduce the precision of the numbers that represent the model’s parameters, which are typically weights and activations.
The two most common quantization cases are float32 -> float16 and float32 -> int8.

This reduction in precision leads to several benefits:

1️⃣ 𝗠𝗼𝗱𝗲𝗹 𝗦𝗶𝘇𝗲 𝗥𝗲𝗱𝘂𝗰𝘁𝗶𝗼𝗻: Lower precision means fewer bits are needed to store each parameter, resulting in a smaller model size.
2️⃣ 𝗙𝗮𝘀𝘁𝗲𝗿 𝗜𝗻𝗳𝗲𝗿𝗲𝗻𝗰𝗲: Lower-precision computations can be performed more quickly, especially on hardware designed for integer operations.
3️⃣ 𝗥𝗲𝗱𝘂𝗰𝗲𝗱 𝗣𝗼𝘄𝗲𝗿 𝗖𝗼𝗻𝘀𝘂𝗺𝗽𝘁𝗶𝗼𝗻: Lower-precision operations consume less power, which is crucial for edge devices like smartphones and IoT sensors.

For example, a deep learning model used for image classification might originally use 32-bit floats for its weights. By applying quantization, these weights can be converted to 8-bit integers. This reduces the model size by a factor of 4 and significantly speeds up inference, making it more efficient to deploy on resource-constrained devices.

In practice, quantization can be applied post-training or during training (quantization-aware training) to minimize any potential loss in model accuracy.

1️⃣ 𝗧𝗼𝗸𝗲𝗻𝗶𝘇𝗮𝘁𝗶𝗼𝗻
Tokenization is where it all begins. The model breaks down text into smaller units called tokens, which could be words or even sub-words. For example, the sentence “Working with NLP is tricky” becomes [“Working”, “with”, “NLP”, “is”, “tricky”, “.”]. This step is crucial because it allows the model to understand input text in a structured way that can be processed numerically.

2️⃣ 𝗦𝘁𝗼𝗽 𝘄𝗼𝗿𝗱 𝗿𝗲𝗺𝗼𝘃𝗮𝗹
Not every word in a sentence carries significant meaning. Stop words like “with” and “is” are common across many sentences but add little to the meaning. By removing these, the model can focus on the more meaningful parts of the text, enhancing efficiency and accuracy.

3️⃣ 𝗟𝗲𝗺𝗺𝗮𝘁𝗶𝘇𝗮𝘁𝗶𝗼𝗻
Lemmatization simplifies words to their base form, making it easier for the model to understand the context without getting bogged down by variations. For instance, words like “talking”, “talked”, and “talk” all get reduced to their root form “talk.

We are then ready for the next step, which is to change the text into a form the computer can understand.

Embeddings & Vectors

Vector embeddings are one of the most fascinating and useful concepts in machine learning. They are central to many NLP, recommendation, and search algorithms.

𝗪𝗵𝗮𝘁 𝗮𝗿𝗲 𝗪𝗼𝗿𝗱 𝗘𝗺𝗯𝗲𝗱𝗱𝗶𝗻𝗴𝘀? 🤔

Word embeddings are a type of representation that allows words with similar meanings to have similar representations. Think of them as vectors in a high-dimensional space where each dimension captures a different aspect of the word’s meaning.

Simply put, words possessing similar meanings or often occuring together in similar contexts, will have a similar vector representation, based on how “close” or “far apart” those words are in their meanings.

𝗔 𝗳𝗮𝗺𝗼𝘂𝘀 𝗲𝘅𝗮𝗺𝗽𝗹𝗲

Consider the equation: “king” – “man” + “women” = “queen”. This example illustrates how word embeddings can capture complex semantic relationships. The vector operations translate semantic similarity as perceived by humans into proximity in a vector space.

𝗕𝗲𝘆𝗼𝗻𝗱 𝗪𝗼𝗿𝗱𝘀: 𝗩𝗲𝗰𝘁𝗼𝗿 𝗘𝗺𝗯𝗲𝗱𝗱𝗶𝗻𝗴𝘀

In other words, when we represent real-world objects and concepts such as images, audio recordings, news articles, user profiles, weather patterns, and political views as vector embeddings, the semantic similarity of these objects and concepts can be quantified by how close they are to each other as points in vector spaces. Vector embedding representations are thus suitable for common machine learning tasks such as clustering, recommendation, and classification.

𝗩𝗲𝗰𝘁𝗼𝗿 𝗗𝗮𝘁𝗮𝗯𝗮𝘀𝗲𝘀

Once you have these embeddings, you need a way to store and query them efficiently. This is where Vector Databases come in. Vector databases are designed to handle high-dimensional data, making them perfect for storing and retrieving word embeddings.

Fine-Tuning vs. Pre-Training

🎯 What is Fine-Tuning?
Think of fine-tuning as the process of specializing in a specific domain, like a college student focusing on medicine. It builds upon the foundation of pre-trained models to adapt them for specific tasks.

🏋️‍♂️ Overcoming “Largeness” Challenges
LLMs are powerful but come with challenges like high computational costs, extensive training time, and the need for vast amounts of high-quality data. Fine-tuning helps overcome these obstacles by:
1️⃣ Reducing computational power requirements
2️⃣ Shortening training time from weeks or months to hours or days
3️⃣ Requiring less data, typically only a few hundred megabytes to a few gigabytes

🔧 Fine-Tuning vs. Pre-Training
While pre-training demands thousands of CPUs and GPUs, fine-tuning can be done with just a single CPU and GPU. Plus, fine-tuning takes significantly less time and data compared to pre-training!

Fine-Tuning & transfer learning

Fine-tuning a pre-trained LLM involves training it on a smaller, task-specific dataset to boost performance. But what happens when labeled data is scarce? Enter zero-shot, few-shot, and multi-shot learning—collectively known as N-shot learning techniques.

𝗧𝗿𝗮𝗻𝘀𝗳𝗲𝗿 𝗟𝗲𝗮𝗿𝗻𝗶𝗻𝗴
These techniques fall under the umbrella of transfer learning. Just like skills from piano lessons can be applied to learning guitar, transfer learning involves leveraging knowledge from one task to enhance performance on a related task. For LLMs, this means fine-tuning on new tasks with varying amounts of task-specific data.

❌💉 Zero-Shot Learning
Zero-shot learning enables LLMs to tackle tasks they haven’t been explicitly trained on. Imagine a child identifying a zebra based on descriptions and knowledge of horses. LLMs use zero-shot learning to generalize knowledge to new situations without needing specific examples.

💉 Few-Shot Learning
Few-shot learning allows models to learn new tasks with minimal examples. Think of a student answering a new exam question based on prior knowledge from lectures. When only one example is used, it’s called one-shot learning.

💉💉💉 Multi-Shot Learning
Multi-shot learning is similar to few-shot learning but requires more examples. It’s like showing a model several pictures of a Golden Retriever to help it recognize the breed and generalize to similar breeds with additional examples.

These techniques make LLM more adaptable and efficient even with limited data 💡

Transformers

𝗪𝗵𝗮𝘁 𝗶𝘀 𝗮 𝗧𝗿𝗮𝗻𝘀𝗳𝗼𝗿𝗺𝗲𝗿?
Introduced in the paper “Attention Is All You Need” 7 years ago, transformers emphasize long-range relationships between words to generate accurate and coherent text.

𝗛𝗼𝘄 𝗧𝗿𝗮𝗻𝘀𝗳𝗼𝗿𝗺𝗲𝗿𝘀 𝗪𝗼𝗿𝗸
Let’s consider an example sentence: “Jane, who lives in New York and works as a software engineer, loves exploring new restaurants in the city.”

1️⃣Text Pre-processing:
The transformer breaks down the sentence into tokens (e.g., “Jane,” “who,” “lives,” etc.) and converts them into numerical form using word embeddings.
2️⃣Positional Encoding:
Adds information about the position of each word in the sequence, helping the model understand the context and relationships between distant words.
3️⃣Encoders:
Use attention mechanisms and neural networks to encode the sentence, focusing on specific words and their relationships.
4️⃣Decoders:
Process the encoded input to generate the final output, such as predicting the next word in the sequence. Why Transformers are Special

𝗟𝗼𝗻𝗴-𝗥𝗮𝗻𝗴𝗲 𝗗𝗲𝗽𝗲𝗻𝗱𝗲𝗻𝗰𝗶𝗲𝘀
Transformers excel at capturing relationships between distant words. For instance, they can understand the connection between “Jane” and “loves exploring new restaurants,” even though these words are far apart in the sentence.

𝗦𝗶𝗺𝘂𝗹𝘁𝗮𝗻𝗲𝗼𝘂𝘀 𝗣𝗿𝗼𝗰𝗲𝘀𝘀𝗶𝗻𝗴
Unlike traditional models that process one word at a time, transformers can handle multiple parts of the input text simultaneously. This speeds up the process of understanding and generating text.

Ever wondered how machines can predict trends, classify images, or make decisions in complex environments?

🤖 What is Machine Learning? ML is a subset of artificial intelligence that enables computers to learn from data without being explicitly programmed. It’s all about making predictions, drawing insights, and identifying patterns to make better decisions.*

There are 3 types of Machine Learning:

1️⃣ Supervised Learning: In supervised learning, we train a model using labeled data, meaning that the target variable (what we want to predict) is known. The model learns from the features (different pieces of information) and labels in the training data to make predictions on new, unseen data.
Example: Predicting heart disease in patients based on factors like age, cholesterol, and smoking habits.

2️⃣ Unsupervised Learning: Unsupervised learning works with unlabeled data, meaning there are no target variables. The model finds patterns and relationships in the data on its own, which is perfect for tasks like anomaly detection and clustering.
Example: Grouping heart disease patients into categories based on feature similarity to research better treatments.

3️⃣ And the least common – Reinforcement Learning: Reinforcement learning is about training models to make sequential decisions, like a robot navigating a path or deciding its next move in a game. The model learns by interacting with its environment and receiving feedback in the form of rewards or penalties. Example: Teaching a computer to play chess by learning from its past moves and their outcomes.

ML Workflow: the main steps

Let’s dive into the four essential steps of the machine learning workflow.
Our Scenario:
Imagine using Berlin apartment sales data to predict future sale prices. With labeled data on square footage, neighborhood, year built, and sale price, this becomes a supervised learning problem.

📊 Step 1 – Extract Features: Begin by reformatting the dataset and selecting relevant features. In our case, we might consider square feet, neighborhood, and distance to the nearest subway station.

⚖️ Step 2 – Split Dataset: Divide the dataset into two parts: training and testing. This separation ensures an accurate evaluation of the model’s performance.

📈 Step 3 – Train Model: Choose a suitable machine learning model and train it using the training dataset. Models range from simple logistic regression to complex neural networks.

🎯 Step 4 – Evaluate: Assess the model’s performance using the test dataset, also known as “unseen data.” Calculate the average error or the percentage of accurate predictions within a certain margin to determine the model’s effectiveness.

If the model meets your performance threshold, it’s ready for use! If not, return to step 3 and fine-tune the model by adjusting its parameters or features.

Deep Learning

Deep learning is a powerful subset of machine learning and takes inspiration from the human brain’s neural networks to solve complex problems!

🔍 What is Deep Learning? Deep learning uses artificial neural networks to learn hierarchical representations of data, enabling AI to process large, unstructured datasets and tackle intricate tasks like computer vision and natural language processing.

🎥 Predicting Box Office Revenue: A Deep Learning Example Imagine using deep learning to predict a movie’s box office revenue based on factors like production budget, advertising, star power, and release timing. A neural network can automatically discover and map relationships between these variables to generate accurate predictions.

🌟 How Does It Work? Neural networks consist of interconnected neurons (or nodes) that process information in layers. In our box office example, neurons might analyze spend, awareness, and distribution to predict revenue. Deep learning networks can have thousands of neurons, enabling them to compute incredibly complex functions and uncover hidden patterns.

When to Choose Deep Learning: Deep learning is ideal for large datasets and complex problems, but it requires powerful computers and more data for training compared to traditional machine learning. When domain knowledge is lacking or the task involves unstructured data, deep learning excels, as neural networks can automatically discover essential features and relationships.

Deep Learning & Computer Vision 📷

Have you ever wondered how self-driving cars navigate or how facial recognition works? Computer vision is the key, and here is a first high-level overview of its core concepts.It enables computers to see and understand the content of digital images, powering applications like self-driving cars, automatic tumor detection, and more.

🎨 Image Data: Images are made up of pixels containing color and intensity information. So, digital images can actually be seen as a bunch of numbers. These numbers can be used as features for your machine learning model.

🧔 Face Recognition: To build a face recognition system, we input images and use a neural network to identify individuals based on pixel intensities. Neurons in the network learn to detect edges, parts of objects, and eventually whole faces, combining this information to output the person’s identity.

🔧 Training the Neural Network: The magic of neural networks lies in their ability to learn from data without explicit programming. By providing images of faces and their corresponding labels, the learning algorithm figures out what each neuron should compute during training.

📱 Applications: Computer vision powers various applications, including facial recognition, self-driving vehicles, automatic tumor detection in CT scans, and even the creation of realistic images, like deepfakes.

Natural Language Processing (NLP): A Deep Dive into Text Analysis 🗣️

NLP enables computers to make sense of text data, powering applications like language translation, chatbots, and sentiment analysis.

Techniques to handle data when it is text:

📜 Bag of Words: When dealing with text data, one simple yet powerful technique is the bag of words. By counting the frequency of important words in each piece of text, we can transform text into numerical features for machine learning models.

🔗 Bag of Words: N-grams: To capture more context, we can count sequences of words, known as n-grams. This technique helps improve the understanding of phrases and sentiment in the text.

❌ Bag of Words: Limitations: Bag of words has limitations, such as not considering synonyms or the context of words. For example, different shades of blue like “sky-blue” and “cerulean” would be treated as separate features.

📦 𝗪𝗼𝗿𝗱 𝗘𝗺𝗯𝗲𝗱𝗱𝗶𝗻𝗴𝘀: One solution to these problems is Word embeddings. They are mathematical representations (vectors) that group similar words together. They also follow intuitive mathematical rules, like “King” – “Man” + “Woman” ≈ “Queen”.

–> 𝘞𝘰𝘳𝘥 𝘌𝘮𝘣𝘦𝘥𝘥𝘪𝘯𝘨𝘴 𝘪𝘴 𝘢 𝘬𝘦𝘺 𝘤𝘰𝘯𝘤𝘦𝘱𝘵 𝘵𝘩𝘢𝘵 𝘥𝘦𝘴𝘦𝘳𝘷𝘦𝘥 𝘵𝘰 𝘣𝘦 𝘥𝘦𝘷𝘦𝘭𝘰𝘱𝘦𝘥 𝘧𝘶𝘳𝘵𝘩𝘦𝘳 𝘪𝘯 𝘢 dedicated article.

📱Applications:

NLP powers various applications, including language translation, chatbots, personal assistants (e.g., Siri and Alexa), sentiment analysis, and more!