🤖 AI/ML Engineer | 🧠 Generative AI & RAG Architect
I am a dedicated AI Engineer specializing in autonomous Agentic workflows, complex RAG architectures, and multimodal ML systems. I focus on engineering zero-hallucination inference pipelines and deploying high-performance machine learning backends for production.
🎓 Currently pursuing my B.Tech in Computer Science Engineering (Class of 2026).
- 🧠 AI & Machine Learning: PyTorch, LangChain, TensorFlow, Llama 3, Hugging Face, Transformers, OpenCV
- ⚙️ Backend & Vector Databases: Python, FastAPI, Qdrant, FAISS, SQL, REST APIs
- 🐳 Infrastructure & DevOps: Docker, Linux, Git, Microsoft Azure
-
🎯 Nexus AI - Agentic Talent Intelligence ATS
- Architected an autonomous AI Agent leveraging advanced RAG architectures to semantically evaluate resumes against job descriptions with zero hallucinations.
- Built a robust, Docker-containerized backend pipeline using FastAPI and Pydantic to statefully manage document data within a Qdrant vector database.
-
🏢 Enterprise HR Policy Assistant
- Engineered a completely local RAG pipeline using Llama 3, LangChain, and FAISS to deliver hallucination-free answers.
- Utilized Server-Sent Events (SSE) via FastAPI to stream real-time LLM inference to client applications, decoupling heavy chunking processes to prevent local OOM crashes.
-
🎭 TriSense - Multimodal Emotion Recognition
- Developed an advanced multimodal pipeline implementing Decision-Level Late Fusion to resolve conflicting signals across video, audio, and text streams.
- Deployed the heavy ML inference engine via a Flask RESTful API, optimizing memory overhead with PyTorch Mixed Precision (Float16).
- 💼 LinkedIn: linkedin.com/in/deepanshu-sharma
- ✉️ Email: deepanshusharma9012@gmail.com