Automation & AI Engineer | Agents • Workflows • Infrastructure | Build. Automate. Scale.
I build autonomous systems. My focus is Agentic AI, automation, and FinTech — designing workflows where software can reason, execute tasks, and scale reliably in real-world financial and data-driven environments. Stack: Python, APIs, Microservices, Linux Specialty: Agentic AI • Automation • FinTech Systems
Engineer's degree, Computer Science
[object Object]
High School Diploma
May 2018 - May 2024
B.E., Computer Science
2024 – 2028 (Expected)
Developed a functional Python server to process and serve global temperature data, leveraging the generative AI capabilities of Anthropic's Claude. I utilized Claude for rapid code generation, real-time debugging of environment and pathing issues, and data format conversion (Text to JSON). This project demonstrates modern AI-assisted development workflows and foundational backend skills.
Developed a personalized movie recommendation system using machine learning techniques within Google Colab. This project focused on suggesting movies based on user preferences and behavior, involving comprehensive data preprocessing, model implementation using Python (Pandas, NumPy, Scikit-learn), robust feature engineering, and rigorous model evaluation. The iterative prototyping demonstrated real-time movie suggestions, significantly enhancing my understanding of recommendation systems, large-scale data handling, and practical machine learning deployment. (Consider adding: "Implemented a [Collaborative Filtering / Content-Based Filtering / Hybrid] approach to generate relevant movie suggestions.")
Architected a production-grade agentic AI system using LangGraph state machines, routing multi-turn conversations across greeting, inquiry, and high-intent lead qualification phases. Engineered a custom RAG pipeline over a JSON knowledge base and deployed a FastAPI webhook integrated with Meta Cloud API and Redis session persistence for stateful WhatsApp conversations.
Built an LLM-powered financial intelligence pipeline using engineered OpenAI API prompts to classify financial news as market-moving or non-market-moving. Designed three prompt strategies and automated CSV ingestion, LLM inference, results export, and analyst-ready statistics across 100+ financial news records.
Trained a regression model on 15,000+ insurance records achieving an R2 of 0.86, and deployed it as a FastAPI REST API sustaining 500+ predictions per minute. Enforced Pydantic schema validation, reducing invalid request volume by 40%.
Delivered a full-stack machine learning system processing 10,000+ rows with a decoupled FastAPI inference backend and Streamlit dashboard, containerized via Docker to reduce setup time by 60%.