Architecting autonomous agentic AI systems, scalable RAG architectures, and statistical machine learning pipelines from raw data to production deployment.
Bridging the gap between theoretical machine learning models and scalable, production-grade AI systems.
Design autonomous multi-agent systems using LangChain and LangGraph capable of search, self-correction, planning, and tool execution to solve multi-step tasks independently.
Implement Retrieval-Augmented Generation architectures combining vector stores, document chunking pipelines, and LLMs to answer domain-specific queries grounded in factual data.
End-to-end data pipelines involving hypothesis testing (Chi-Square, Pearson), feature engineering, statistical validation, and distinguishing correlation from causation.
I am an early-career Data Scientist & AI Systems Engineer with a strong academic background in AI and Data Science. My focus spans building multi-agent workflows, fine-tuning NLP classification models, and developing end-to-end machine learning solutions.
Fluent in Python and SQL with a solid foundation in statistical inference, I specialize in taking raw, unstructured web data and transforming it into user-facing web applications deployed on cloud environments.
Direct links to real-world AI repositories and analytics systems.
Autonomous multi-agent research system orchestrating four agents (search, reader, writer, critic) built with LangChain & LangGraph. Scrapes web sources via Tavily/BS4 and generates self-critiqued reports streamed live on Streamlit.
AI application extracting facial emotion features using DeepFace CNNs and OpenCV. Dynamically generates personalized YouTube music playlists in under 5 seconds via YouTube Data API v3 integration.
Performed end-to-end data analysis on a 9,994-row retail dataset using SQLite, Python, and Power BI. Wrote advanced SQL window functions uncovering that 18.7% of orders were loss-making and discounts >40% generated negative profit.
End-to-end statistical modeling pipeline utilizing Pearson Correlation and Chi-Square hypothesis testing to statistically validate feature selection, pinpointing smoking status as the primary cost driver.
Open to AI Engineering, Data Science, and Machine Learning roles.