Hey there! I'm thrilled you stopped by my personal repo. Whether you're here to explore some cool AI projects, collaborate, or just check out what I’ve been working on, I’m excited to share it with you!
I have a mix of software engineering and strong math skills. As a full-stack AI practitioner at @Microsoft, I integrate advanced theory with practical application, and I work hard crafting AI/ML solutions at scale for Fortune 500 giants with a particular focus on Generative AI and LLMOps.
- Complex Agent-Human Interaction: Diving deep into the intricate dynamics of how automated agents and humans interact, aiming to enhance the synergy between users and AI-driven systems for optimized outcomes.
- Multi-Agent Architectures: Designing and implementing multi-agent architectures to solve real-world problems, focusing on the integration and coordination of various agents. These architectures are being tailored for practical application in industries and are being integrated into the market alongside small language models (SLMs) for domain-specific knowledge.
- Reinforcement Learning Techniques: Actively exploring and mastering various reinforcement learning strategies, particularly through DPO, focusing on fine-tuning SLMs that can adapt and excel in a myriad of environments and domain-specific scenarios.
- Large Language Models (LLM) Evaluation: Engaged in the systematic evaluation of Large Language Models, understanding their strengths, limitations, and potential, with particular emphasis on real-world applications and implications.
- Latent Diffusion Algorithms: Delving into the intricacies of latent diffusion algorithms, studying their efficiency, scalability, and performance, with an eye toward harnessing these powerful tools for robust and reliable AI solutions.
- Established and led an MLOps practice, deploying both real-time and batch ML solutions globally at Concentrix consulting services.
- Principal author and designer of the MLOps accelerator framework at Levi's & Co Tech.
- Co-authored, developed, and engineered a set of Python SDKs at Levi's & Co Tech, simplifying ML model integration into software for large DS organizations.
- Designed a state-of-the-art product classification and embedding generation system using advanced Computer Vision algorithms.
- Led multiple algorithms to production-grade ML systems within Fortune 500 enterprises during my consulting stage at Concentrix, from concept to deployment.
Check out my blog: Pablo's AI Corner
-
Modernize Your Document Management with Azure AI
Learn how to enhance OCR, intelligent tagging, and NER with Azure AI. -
Key Technical Challenges in GenAI Production
Explore key hurdles when transitioning Generative AI applications to production. -
Effective LLM/SLM Evaluation Framework
Discover strategies for evaluating large and small language models in enterprise settings.
I’m looking to collaborate on Open-source challenges, innovative product ideas, and partnerships with bright AI/ML practicioners.
- 💌 Shoot me an email: [email protected]
- 🔗 LinkedIn: Let’s get linked!