• Developed and optimized CI/CD workflows with GitHub Actions and custom shell scripts, achieving end-to-end automated deployment and ensuring daily status reports were shared via email for enhanced team transparency.
• Built an AI-driven client QnA framework using LLMs with CrewAI guardrails for safe, compliant, and reliable conversations.
• Made data pipelines for Discord and YFinance integration, enabling Retrieval-Augmented Generation (RAG) Chat to generate a brain by feeding structured information into the model’s knowledge base, along with two additional pipelines to automate data flow and knowledge merging
• Spearheading the development of a BLE-based Indoor Positioning System (IPS), achieving up to 87% room- level accuracy using machine learning and RSSI signal strength analysis.
• Processed and modeled 200,000+ RSSI signals using Python (Scikit-learn, Pandas, Jupyter), leveraging Random Forest and KNN algorithms to get high-performing classification and regression models.
• Established a beacon trajectory tracking module that mapped beacon movement in real time, improving the robustness of the IPS and demonstrating reliable position estimates across diverse room layouts.
Restricted contact info. You need company account to have access.
