const naman = {
role: "AI Engineer | Full Stack Developer | Researcher",
past_roles: {
intern: "AI Developer @ ThinkingStone — Voice AI & Autonomous Agents",
},
education: "B.Tech @ NIT Rourkela (2024-2028)",
Portfolio & resume : { namanjaiswal.in }
expertise: {
ai_ml: [
"🧠 Deep Learning & Transformers (PyTorch, TensorFlow)",
"🤖 Multi-Agent Systems (LangChain, CrewAI, AutoGen, LangGraph)",
"🗣️ Voice AI & Real-time Systems (LiveKit, WebRTC, Deepgram)",
"📊 RAG Pipelines (ChromaDB, Pinecone, Weaviate, FAISS)",
"🔧 LLM Fine-tuning & Prompt Engineering",
"🎯 Production ML, MLOps & LLMOps"
],
engineering: [
"💻 Full-Stack (FastAPI, Flask, React, Next.js, Streamlit)",
"🐳 DevOps (Docker, Kubernetes, GCP, Fly.io, CI/CD)",
"🗄️ Databases (PostgreSQL, MongoDB, Supabase, Redis, Vector DBs)",
"🔌 Real-time (WebSockets, RPC Workflows, MCP Framework)"
],
research: [
"🔬 NLP & Transformer Architectures",
"👁️ Computer Vision & Vision Transformers",
"🎙️ Conversational AI & Human-Machine Interaction",
"🤖 Autonomous Agent Design & Cognitive Architectures"
]
},
achievements: {
ntse: "🏆 NTSE Scholar — Top 1000 out of 1M+ students nationwide",
child_scientist: "🔬 State Child Scientist — Govt. of India",
school: "📚 96.2% in CBSE Class 10th"
},
philosophy: "Take AI from research papers → production systems"
};| Project | Impact |
|---|---|
| 🎙️ RAG Voice AI | Real-time voice agent, sub-100ms latency |
| 📡 VoiceMon | Production observability for Voice AI |
| 📄 CogniDocs | RAG document intelligence platform |
| ⚡ DistriNews | Distributed Transformer training with DDP |
| 🛒 EcomIntelligence | RAG + MCP product assistant |
| 🧠 GPT from Scratch | 124M params, 94% spam accuracy |
| 👁️ ViT Language Model | Vision Transformer + decoder LM |
| 🏥 VetScribe AI | Medical transcription, 2-3x real-time |
| 🏢 ThinkingStone | 99.9% uptime, zero-downtime deploys |
| 🤖 Agents | 20+ autonomous agents built |
June 2025 – August 2025 · Remote
🎙️ Voice AI & Autonomous Agent Systems
- Architected core AI voice agent logic — intent parsing, dialogue state management, intelligent fallback strategies
- Built autonomous AI agents with tool-use capabilities using ReAct pattern for reasoning & action selection
- Developed real-time audio pipelines with LiveKit SFU for dynamic multi-party voice sessions & agent handoffs
- Created microservices using Docker + MCP framework for call orchestration, session metadata, and RPC workflows
- Deployed on Fly.io edge network with auto-scaling, health checks, and zero-downtime rolling updates → 99.9% uptime
- Integrated Supabase + WebSockets for live ingestion, transcript storage, and agent memory persistence → sub-100ms latency
🎙️ Voice AI & Real-time Systems
LiveKit SFU · WebRTC · Deepgram STT · ElevenLabs TTS · Vosk · Audio Pipeline Development · Intent Parsing · Dialogue State Management · WebSockets · Real-time Streaming · RPC Workflows
🤖 Agentic AI Deep Dive
LangChain Agents · ReAct Pattern · Tool Use & Function Calling · Multi-Agent Orchestration · Agent Memory Systems · Task Planning & Decomposition · Chain-of-Thought Prompting · Tree-of-Thoughts · Agent Evaluation Frameworks · CrewAI · AutoGen · MetaGPT · LlamaIndex Agents · DSPy · OpenAI Assistants API · Agent Protocol · Cognitive Architectures
🏛️ National Institute of Technology (NIT) Rourkela
Aug 2024 – May 2028
B.Tech — Expected Graduation: 2028
✅ Complete Data Science, ML, DL, NLP Bootcamp — Udemy
✅ Generative AI with LangChain & HuggingFace — Udemy
✅ Mathematics for Data Science & GenAI — Udemy
📚 Relevant AI/ML Coursework
| Domain | Courses |
|---|---|
| Core AI/ML | Machine Learning, Deep Learning, RL, Statistical Learning Theory |
| Deep Learning | Neural Networks, CNNs, RNNs, GANs, VAEs |
| NLP & LLMs | NLP, LLMs & Transformers, Sequence Models, Text Mining |
| Agentic AI | Multi-Agent Systems, Autonomous Agents, Knowledge Representation |
| Computer Vision | CV, Image Processing, Object Detection, Visual Learning |
| Advanced | RL for Robotics, PGMs, Federated Learning, MLOps, AI Ethics |
| Foundations | DSA, Linear Algebra, Probability & Statistics, Distributed Systems |
💬 I'm always open to collaborating on AI/ML research, open-source contributions, and engineering challenges at scale.
If you're working on LLMs, multi-agent systems, voice AI, or production ML — let's connect!
💡 "Building production-grade AI systems that actually ship — not just pretty demos."
⭐️ From Naman Jaiswal · Last Updated: Feb 2026 🚀