Headline

Data Science & Machine Learning Engineer - Acuity Analytics | Building Enterprise GenAI & RAG Systems | Ex-Microsoft Intern | LLMs, NLP, ML & AI Infrastructure

About

AI Engineer specializing in Generative AI, Retrieval-Augmented Generation (RAG), LLM Orchestration, and Agentic AI systems - building scalable, production-grade GenAI infrastructure for enterprise applications. Currently at Acuity Analytics, I work on Acuity Agent Fleet and cloud-agnostic RAG infrastructure, designing systems that improve knowledge retrieval, flexibility, and AI performance at scale. My work spans agentic architectures, custom retrieval pipelines, and LLM orchestration frameworks like LangGraph and LangChain - moving GenAI from experimentation into real production deployment. Previously, as an AI/ML Engineering Intern at Microsoft, I worked on an AI-driven taxonomy mapping initiative for the global L&D team - fine-tuning GPT and BERT-based models alongside Random Forest classifiers to improve taxonomy mapping accuracy by 20%. Core focus areas: 🔹 Generative AI & LLM Applications 🔹 Agentic RAG & LLM Orchestration (LangGraph, LangChain, MCP) 🔹 NLP, Text Embeddings & Vector Search 🔹 AI Infrastructure, Automation & MLOps 🔹 Machine Learning & Deep Learning (PyTorch, TensorFlow, Transformers) 🔹 Cloud-Native AI Systems (AWS, GCP, Azure OpenAI) Tech Stack: Python, Java, JavaScript (React/Node) | PyTorch, TensorFlow, Scikit-learn, Hugging Face | LangChain, LangGraph, RAGFlow, MLFlow | AWS (SageMaker, Bedrock, ECS, S3, Step Functions), GCP, Azure OpenAI | Docker, Podman, Microservices, Temporal | PostgreSQL, MongoDB, Vector DBs (OpenSearch/PGVector) Open to collaborating on AI products, applied research, and Generative AI systems that solve real enterprise problems - always happy to connect with fellow builders, researchers, and recruiters in the AI/ML space.

Location

Bengaluru

AI stack

claude code

cursor

gemini

langchain

mcp

Work experience

Data Science and Machine Learning Engineer at Acuity Analytics, Sep 2025 - Present · 1 yr, Bengaluru, Architected and self-hosted a cloud-agnostic RAG knowledge base on AWS ECS (built on and heavily extended an open-source connector framework), supporting 40+ enterprise data sources (Salesforce, SharePoint, S3, Zoho CRM, Todoist, etc.) — filling a gap unaddressed by managed offerings like AWS Knowledge Bases, which lack native support for several of these connectors Designed and built new connectors (including S3) and added OpenSearch and PgVector as pluggable vector store backends, enabling flexible, cost-optimized vector search across deployments Engineered a custom chunking strategy with metadata enrichment and table-of-contents-aware ingestion, improving retrieval precision for downstream agentic workflows Reduced infrastructure cost and improved query latency versus managed cloud alternatives by self-hosting the full RAG pipeline (parsing, chunking, embedding, retrieval) Integrated the platform as the retrieval backbone (via MCP-based data connectors) for two production AI products: AgentFleet Assistant (agentic AI assistant for private-markets workflows) and FolioSure (AI-powered loan operations knowledge assistant)

Intern Data Scientist - Research and Development at Acuity Analytics, Mar 2025 - Aug 2025 · 6 mos, Bengaluru, Karnataka, India, Worked with the Business Excellence and Automation Team (BEAT) - Data Science - Research and Development Team on various techniques to apply a multi-class classification to identify financial attributes from financial statements of different companies. The internship involved extensive research and experimentation about different Object Character Recognition (OCR) Techniques, Text Embeddings and Large Language Model results.

Intern at Microsoft, Jun 2024 - Aug 2024 · 3 mos, Bengaluru, - Contributed to an enterprise AI taxonomy mapping initiative for Microsoft’s global Learning & Development organization. - Fine-tuned GPT and BERT-based NLP models to improve hierarchical taxonomy mapping accuracy by 20%. - Applied Random Forest classification techniques to optimize prediction quality across multi-level skill taxonomies. - Experimented with multiple NLP and embedding approaches to improve classification reliability and scalability. - Worked closely with cross-functional stakeholders within Microsoft Delivery Excellence to support AI-driven operational efficiency initiatives.

Education

Bachelor of Technology, Computer Science Engineering (Specialization in Artificial Intelligence and Machine Learning), Graphic Era Deemed to be University

Foundation degree, Programming and Data Science, Indian Institute of Technology, Madras