Building Intelligent Systems at Enterprise Scale
Agentic AI • LLMs • RAG • Context Engineering • Enterprise AI • AI Safety
I'm a Data Scientist at Endress+Hauser, where I architect and deploy enterprise-grade AI solutions — from agentic AI systems that automate purchase order workflows with 90% efficiency gains, to pioneering one of the earliest production RAG deployments in industry.
With a Master's in AI & Data Science and experience spanning PwC, Noviter, and PTCL, I bring deep expertise in LLMs, context engineering, agentic workflows, and AI safety. I'm also a Coursera instructor, active conference speaker, and hackathon champion.
“HITL 2.0: Designing Cooperative Workflows with Agentic AI” — practical use cases and design patterns for agentic workflows with human-in-the-loop control.
“Beyond the Buzz: Practical Applications of LLM Agents” — showcasing tangible use cases, prompt engineering, domain adaptation, and real-world challenges.
“Optimizing LLMs for E+H” — insights on deploying LLMs at enterprise scale and building a successful production RAG on internal knowledge bases.
“LLMs for Dashboarding” — demonstrated how conversational AI translates user queries into SQL, revolutionizing data exploration and dashboard automation.
“The Future of AI: Democratizing AI Across Disciplines” — emerging trends, tools, and resources for making AI accessible and fostering interdisciplinary collaboration.
“LLMs: From Inception to Future Directions” — the evolution of LLMs, highlighting significant breakthroughs and predictions for future advancements.
Spoke about harnessing Generative AI for transformative innovation across industries.
Anomaly detection via autoencoders and time series demand forecasting for industrial applications.
Winner of BaselHack 2024, building a scalable RAG with web browser and memory agents for MDPI.
Instructor for Data Science and Machine Learning Guided Projects Lab on Coursera.
Invited speaker at prestigious international conferences including ML Week Europe, Basel DataScience, European Chatbot Summit, and more.
Published research in IEEE on AI and machine learning applications.
Human-in-the-loop, multi-agent AI framework processing 300–1,000+ page user manuals to generate high-quality how-to guides and knowledge articles, with human oversight of final content.
Multi-agent feedback analysis workflow for survey responses — automated reading, summarization, sentiment analysis, theme identification, and structured insight-rich reports.
Multi-agent content generation system — agents gather trends from news/social platforms, build plans, and generate tailored posts and visuals for different channels with human approval.
Developed XAI module for PwC to explain how drivers impact ML decisions, improving customer trust. Deployed AI/ML solutions via Palantir Foundry for time series and forecasting.
MySQL-based recommendation engine built with Streamlit on the MovieLens 25M dataset, including analytical reporting.
View on GitHubImplementation of the Boruta algorithm for robust feature selection using entropy and information gain on high-dimensional datasets.
View on GitHubInterested in collaboration, speaking engagements, or just want to chat about AI? Feel free to reach out.