Motivated AI/ML Engineer specializing in developing rapid Proof-of-Concepts (PoCs) and Minimum Viable Products (MVPs) for Machine Learning and Generative AI solutions. Proficient in engineering LLM/RAG enterprise architectures, architecting end-to-end ML workflows, mitigating hallucination & bias, and optimizing production data pipelines with Python and SQL. Experienced in executing mathematical SME operations and full-stack model evaluations.
- Architected and deployed scalable machine learning workflows and automated 6-phase ETL pipelines for time-series data reporting.
- Engineered predictive analytics pipelines to drive actionable operational business intelligence.
- Optimized production inference pipelines to improve processing latency and ensure high scalability.
PDF Chatbot - Enterprise RAG System
Constructed a live RAG pipeline utilizing LangChain (LCEL) and OpenAI LLMs. Implemented hallucination monitoring and deployed via Gradio UI on Hugging Face Spaces.
LangChain OpenAI GradioShopWhisper AI - Multimodal Chatbot
Integrated OpenAI Whisper, Chat Completions, and gTTS in Python. Optimized pipeline latency by executing EDA on live audio-text feeds.
Whisper AI Multimodal gTTSEnterprise Customer Churn Predictor
Executed end-to-end ML model using XGBoost and Random Forest. Achieved high Precision/Recall/ROC-AUC scores with automated CSV batch predictions and an interactive web interface.
XGBoost Scikit-Learn EDAPrivate Local Offline RAG System
Built an air-gapped secure RAG architecture using local LLMs (Ollama) and ChromaDB vector databases for local document retrieval, with feedback loops to refine prompt context.
Ollama ChromaDB Offline AISales Operations Analytics & Enterprise ETL Pipeline
Worked with relational databases to collect, clean, and preprocess transaction data into star schema structures in MySQL, feeding downstream Power BI reporting.
MySQL Power BI ETL