QUANT AI Lab

AI is becoming deeply embedded in financial decision-making.  Credit scoring. Fraud detection. Portfolio optimization. 𝗕𝘂𝘁 𝗮𝘀 𝗔𝗜 𝘀𝘆𝘀𝘁𝗲𝗺𝘀 𝗯𝗲𝗰𝗼𝗺𝗲 𝗺𝗼𝗿𝗲 𝗽𝗼𝘄𝗲𝗿𝗳𝘂𝗹, 𝗼𝗻𝗲 𝗾𝘂𝗲𝘀𝘁𝗶𝗼𝗻 𝗺𝗮𝘁𝘁𝗲𝗿𝘀 𝗺𝗼𝗿𝗲 𝘁𝗵𝗮𝗻 𝗲𝘃𝗲𝗿: 𝗖𝗮𝗻 𝘄𝗲 𝗮𝗰𝘁𝘂𝗮𝗹𝗹𝘆 𝗲𝘅𝗽𝗹𝗮𝗶𝗻 𝘁𝗵𝗲𝗶𝗿 𝗱𝗲𝗰𝗶𝘀𝗶𝗼𝗻𝘀? Because in finance, performance alone is not enough. When AI operates as a black box, the risks go beyond technical limitations: compliance exposure, biased outcomes, reputational damage, loss of trust. And trust is not optional in financial services. 𝗧𝗵𝗶𝘀 𝗶𝘀 𝘄𝗵𝗲𝗿𝗲 𝗘𝘅𝗽𝗹𝗮𝗶𝗻𝗮𝗯𝗹𝗲 𝗔𝗜 (𝗫𝗔𝗜) 𝗯𝗲𝗰𝗼𝗺𝗲𝘀 𝗰𝗿𝗶𝘁𝗶𝗰𝗮𝗹.  By making AI systems more transparent, interpretable, and auditable, financial institutions can better understand outcomes, strengthen governance, and align with evolving regulatory expectations. At QUANT AI Lab, we believe the future of financial AI is not only intelligent. 𝗜𝘁 𝗶𝘀 𝗲𝘅𝗽𝗹𝗮𝗶𝗻𝗮𝗯𝗹𝗲, 𝘁𝗿𝘂𝘀𝘁𝘄𝗼𝗿𝘁𝗵𝘆, 𝗮𝗻𝗱 𝗮𝗰𝘁𝗶𝗼𝗻𝗮𝗯𝗹𝗲. #ArtificialIntelligence #ExplainableAI #XAI #Finance #FinTech #RiskManagement #AI #MachineLearning #EUAIAct 

When an AI researcher says “Let me simplify this”… Spoiler: things get complicated real fast 😅 Lilian accepted the challenge 👇 Want the full breakdown? You’ll find Lilian’s article in our newsletter dropping Monday 👀

Many AI projects don’t fail during development.  𝗧𝗵𝗲𝘆 𝗳𝗮𝗶𝗹 𝗮𝗳𝘁𝗲𝗿 𝗱𝗲𝗽𝗹𝗼𝘆𝗺𝗲𝗻𝘁. Because building a high-performing model is only part of the equation. Production introduces a different level of complexity:  • data changes over time  • real-world performance issues  • inconsistencies between testing and production  • unpredictable user behavior Accuracy alone is not enough.  What creates value is an AI system that remains reliable over time and performs under real-world conditions. 𝗧𝗵𝗲 𝗿𝗲𝗮𝗹 𝗰𝗵𝗮𝗹𝗹𝗲𝗻𝗴𝗲 𝗶𝘀 𝗻𝗼𝘁 𝗯𝘂𝗶𝗹𝗱𝗶𝗻𝗴 𝘁𝗵𝗲 𝗺𝗼𝗱𝗲𝗹. 𝗜𝘁 𝗶𝘀 𝗺𝗮𝗸𝗶𝗻𝗴 𝗶𝘁 𝘄𝗼𝗿𝗸 𝗰𝗼𝗻𝘀𝗶𝘀𝘁𝗲𝗻𝘁𝗹𝘆 𝗶𝗻 𝗽𝗿𝗼𝗱𝘂𝗰𝘁𝗶𝗼𝗻. At QUANT AI Lab, we believe the real value of AI starts when systems perform reliably in production. Swipe through to explore some of the most common reasons AI models fail once deployed. #ArtificialIntelligence #MachineLearning #MLOps #AIEngineering #DataScience #QUANTAILab 

𝗪𝗲 𝗮𝗿𝗲 𝗱𝗲𝗹𝗶𝗴𝗵𝘁𝗲𝗱 𝘁𝗼 𝘄𝗲𝗹𝗰𝗼𝗺𝗲 José Manuel Ordóñez Suárez 𝘁𝗼 QUANT AI Lab 𝗮𝘀 𝗗𝗶𝗿𝗲𝗰𝘁𝗼𝗿 𝗼𝗳 𝗤𝗨𝗔𝗡𝗧 𝗚𝗮𝗹𝗶𝗰𝗶𝗮, an appointment that reflects the company’s commitment to bringing on board leaders with a strong executive track record and deep expertise in the technology sector. José Manuel Ordóñez Suárez brings more than 30 years of experience in the telecommunications industry, having held executive leadership roles in both Spain and Portugal. Throughout his career, he has led large-scale teams and projects, providing him with a strong strategic vision and proven management capabilities in highly demanding environments. QUANT AI Lab warmly welcomes José Manuel Ordóñez Suárez and wishes him every success in this new role leading Quant Galicia.

𝗪𝗵𝗮𝘁 𝗱𝗼 𝘁𝗵𝗲 𝗥𝗼𝗹𝗹𝗶𝗻𝗴 𝗦𝘁𝗼𝗻𝗲𝘀 𝗵𝗮𝘃𝗲 𝘁𝗼 𝗱𝗼 𝘄𝗶𝘁𝗵 𝗔𝗜? More than you might think. One of the biggest misconceptions around artificial intelligence is that better outcomes always require more data, bigger models, and more computing power. But what if useful AI was actually about relevance? In his latest interview with Les Echos Investir, our CEO Bertrand K. Hassani, PhD shares a perspective that challenges some common assumptions about AI: • Why bigger models are not always better  • Why more data does not necessarily mean better outcomes  • How AI learns and where its real limitations lie  • Why we should remain critical of AI-generated answers And much more. A thoughtful conversation about AI beyond the hype: pragmatic, nuanced, and grounded in real-world applications. 𝗥𝗲𝗮𝗱 𝘁𝗵𝗲 𝗳𝘂𝗹𝗹 𝗶𝗻𝘁𝗲𝗿𝘃𝗶𝗲𝘄 𝗶𝗻 𝗟𝗲𝘀 𝗘𝗰𝗵𝗼𝘀 Investir : https://lnkd.in/eP2JEfy2 QUANT AI Lab Bertrand K. Hassani, PhD Manuel Revuelta Julien Doutremépuich Laurent LUCIEN, Ph.D.

𝗔𝗜 𝗱𝗼𝗲𝘀𝗻’𝘁 𝗰𝗿𝗲𝗮𝘁𝗲 𝘃𝗮𝗹𝘂𝗲 𝗶𝗻 𝘁𝗲𝘀𝘁𝗶𝗻𝗴 𝗲𝗻𝘃𝗶𝗿𝗼𝗻𝗺𝗲𝗻𝘁𝘀.  𝗜𝘁 𝗰𝗿𝗲𝗮𝘁𝗲𝘀 𝘃𝗮𝗹𝘂𝗲 𝗶𝗻 𝗽𝗿𝗼𝗱𝘂𝗰𝘁𝗶𝗼𝗻.  Yet many projects are still evaluated in conditions far from operational reality. In computer vision especially, the gap between testing performance and real-world performance is often underestimated. Systems may look impressive in demos, yet struggle to deliver real operational value once deployed. At QUANT AI Lab, we see one recurring pattern:  𝗥𝗲𝗮𝗹 𝗽𝗲𝗿𝗳𝗼𝗿𝗺𝗮𝗻𝗰𝗲 𝗱𝗲𝗽𝗲𝗻𝗱𝘀 𝗹𝗲𝘀𝘀 𝗼𝗻 𝗽𝗲𝗿𝗳𝗲𝗰𝘁 𝗰𝗼𝗻𝗱𝗶𝘁𝗶𝗼𝗻𝘀, 𝗮𝗻𝗱 𝗺𝗼𝗿𝗲 𝗼𝗻 𝗵𝗼𝘄 𝘄𝗲𝗹𝗹 𝗔𝗜 𝗵𝗮𝗻𝗱𝗹𝗲𝘀 𝗶𝗺𝗽𝗲𝗿𝗳𝗲𝗰𝘁 𝗼𝗻𝗲𝘀. Because accuracy only matters if it holds up in production. #ComputerVision #ArtificialIntelligence #EnterpriseAI #IndustrialAI #MachineLearning  

Everyone wants better AI. Very few companies ask a simpler question: 𝗖𝗮𝗻 𝘄𝗲 𝗮𝗰𝘁𝘂𝗮𝗹𝗹𝘆 𝘁𝗿𝘂𝘀𝘁 𝘁𝗵𝗲 𝗱𝗮𝘁𝗮 𝗯𝗲𝗵𝗶𝗻𝗱 𝗶𝘁? Poor data quality costs companies an average of $𝟭𝟮.𝟵𝗠 𝗲𝘃𝗲𝗿𝘆 𝘆𝗲𝗮𝗿, not because models fail, but because decisions do. The real competitive advantage isn’t better AI. It’s better data. At QUANT AI Lab, we help companies make AI and business decisions more reliable by improving how data is monitored, trusted and used. #ArtificialIntelligence #DataQuality #BusinessPerformance #DataDriven #AI #DecisionMaking

📣𝐅𝐨𝐐𝐮𝐬 𝐀𝐩𝐫𝐢𝐥 𝟐𝟎𝟐𝟔: 𝐓𝐡𝐞 𝐐𝐔𝐀𝐍𝐓 𝐀𝐈 𝐋𝐚𝐛 𝐍𝐞𝐰𝐬𝐥𝐞𝐭𝐭𝐞𝐫 𝐢𝐬 𝐋𝐈𝐕𝐄! This month’s edition explores a central question for enterprise AI: how can organizations move from performance to control, while keeping decision-making transparent and aligned with real-world constraints? 𝐄𝐝𝐢𝐭𝐨: 𝐀𝐈 𝐚𝐧𝐝 𝐭𝐡𝐞 𝐪𝐮𝐞𝐬𝐭𝐢𝐨𝐧 𝐨𝐟 𝐜𝐨𝐧𝐭𝐫𝐨𝐥 In this month’s editorial, Julien Doutremépuich examines a key shift in enterprise AI: the challenge is no longer just performance, but control. As AI becomes embedded in critical processes, the real question becomes what is being delegated, and what remains under control. 𝐅𝐨𝐜𝐮𝐬 𝐩𝐫𝐨𝐝𝐮𝐜𝐭: 𝐞-𝐏𝐨𝐫𝐭𝐟𝐨𝐥𝐢𝐨 We present e-Portfolio, our portfolio optimization solution designed to structure and enhance decision-making in complex financial environments, enabling more consistent, traceable and forward-looking allocation strategies. 𝐀𝐈 𝐠𝐨𝐯𝐞𝐫𝐧𝐚𝐧𝐜𝐞: 𝐭𝐫𝐚𝐧𝐬𝐩𝐚𝐫𝐞𝐧𝐜𝐲 𝐚𝐧𝐝 𝐡𝐮𝐦𝐚𝐧 𝐨𝐯𝐞𝐫𝐬𝐢𝐠𝐡𝐭 We highlight Julien Doutremépuich’s viewpoint on the urgency of transparent and human AI governance, where efficiency cannot replace responsibility, traceability and human judgment. Read the full article: https://lnkd.in/eByyDh_d 𝐀𝐈 𝐚𝐧𝐝 𝐭𝐡𝐞 𝐂𝐅𝐎 𝐟𝐮𝐧𝐜𝐭𝐢𝐨𝐧 Following a DAF Mag feature with Bertrand K. Hassani, PhD and Julien Doutremépuich, we explore how AI is transforming the finance function, moving beyond automation toward more analytical, forward-looking and controlled decision processes. Read the full article: https://lnkd.in/dttUx4jD 𝐄𝐱𝐩𝐞𝐫𝐭 𝐩𝐞𝐫𝐬𝐩𝐞𝐜𝐭𝐢𝐯𝐞: 𝐄𝐮𝐫𝐨𝐩𝐞 𝐚𝐧𝐝 𝐀𝐈 Bertrand K. Hassani, PhD shares a perspective on Europe’s AI trajectory, highlighting how regulation can also support a model based on trust, governance and long-term value. Read the full article: https://lnkd.in/eqBbdvuP 𝐖𝐡𝐨 𝐚𝐫𝐞 𝐰𝐞? Finally, this edition presents QUANT AI Lab’s positioning and expertise in applied AI, quantitative finance and decision intelligence for complex business environments. #AI #AIGovernance #DecisionMaking #FinancialServices #AppliedAI #CFO #PortfolioOptimization #QUANTAILab #FoQusNewsletter

What if Europe’s position in AI is being misunderstood? At QUANT AI Lab, this question is worth asking. Glad to see our CEO Bertrand K. Hassani, PhD featured in Channel Partner (BPS). The conversation is usually framed around speed and scale. But that framing misses something important. In Europe, the focus is also on how AI is governed, deployed, and actually used in real-world environments. That may look slower. But it addresses a different challenge: turning AI performance into real adoption. Because what matters is not just what AI can do, but what organizations can trust and sustain over time. Different priorities. Different trajectory.

For finance leaders, the real challenge is not adopting AI, it’s generating tangible value from it. At QUANT AI Lab, we believe this value doesn’t come from generic, one‑size‑fits‑all approaches, but from solutions designed to adapt to business realities, data, and the specific challenges of each organization. In a recent article published by DAFMAG.FR, Bertrand K. Hassani, PhD, CEO of QUANT AI Lab, and Julien Doutremépuich, France Director, share this conviction: Moving beyond standard tools to solutions that genuinely help finance teams anticipate and make better decisions. For finance organizations, the goal is not automation alone, but the ability to:  - Anticipate expenses more accurately  - Detect early warning signals  - Improve financial steering and performance 🔗 Read the article: https://lnkd.in/dttUx4jD