I'm a Data Scientist specializing in AI systems: applying rigorous statistical thinking, evaluation methodology, and experimentation to build Generative AI applications that deliver measurable business impact in production.

My foundation is quantitative rigor: experimental design, metrics that matter, statistical significance, and systematic evaluation. I bring this discipline to AI engineering: architecting multi-agent systems, building RAG pipelines, and optimizing LLMs so they perform predictably at scale.

• 🎓 MS in Applied Data Science - University of Florida (specialization: AI/ML)
• 📊 Data Science expertise: experimental design, statistical evaluation, metrics & monitoring, model performance analysis
• 🤖 AI Systems builder: LLM architecture, multi-agent orchestration, RAG systems, prompt engineering & optimization
• ⚙️ Full-stack practitioner: system design → hypothesis → evaluation → fine-tuning → deployment → production observability

Core expertise: experimental design, model evaluation, metrics & monitoring, interpretability

Both traditional ML and deep learning expertise: supervised/unsupervised learning, neural networks, model architecture

Building rigorous evaluation workflows: metrics design, statistical testing, observability

Applying data science rigor to LLMs: agentic architectures, RAG, prompt optimization, eval

Building interactive data apps & interfaces for AI systems

End-to-end Gen AI system that automates enterprise approval workflows via natural language (text, voice, image). Integrates Retrieval-Augmented Generation against policy documents stored in Cloudflare Vectorize, routes requests intelligently through an agentic decision layer, and auto-approves routine items — reducing manual overhead at scale.

Key concepts: RAG · Agentic Routing · Multi-modal Input · Vector Search · Serverless AI

A fully orchestrated multi-agent system simulating an AI-driven software engineering team. Agents are assigned distinct roles — architect, developer, reviewer — and collaborate autonomously via the Claude Agents SDK, demonstrating task decomposition, inter-agent messaging, and goal-driven execution.

Key concepts: Multi-Agent Orchestration · Claude SDK · Task Delegation · Autonomous Agents

AI-powered implementation of the classic Bulls & Cows guessing game with intelligent move generation and natural language interaction.

Key concepts: LLM Integration · Game AI · Interactive UX

▸ Agentic AI Systems        ████████████░░   Shipping production-grade multi-agent pipelines
▸ RAG & Vector Search       ███████████░░░   Grounding LLMs on real-world enterprise data
▸ LLM Evaluation & Evals    █████████░░░░░   Building robust eval + observability frameworks
▸ AI Engineering (Infra)    ████████░░░░░░   Inference optimization, deployment, monitoring

I'm Actively looking for Full Time Data Science/AI Engineering/Software Engineering roles, and research collaborations in cutting edge technologies.

📬 Connect on LinkedIn
📧 Email me