Anshu tiwari

Anshu tiwari

AI Engineer | Generative AI · Agentic Systems · RAG · Reinforcement Learning

Agartala, India

#OpenToWork

About

AI/ML engineer specializing in agentic systems and applied reinforcement learning, with hands-on experience architecting multi-agent LangGraph pipelines, RAG systems, and RL-driven decision engines deployed with production-grade rigor (HITL authorization, deterministic policy enforcement, audit trails). Solved 700+ DSA problems; ranked top 4 percentile among 1.6M+ JEE Main candidates. Currently building AI-governed control systems combining LLM planning with reinforcement learning execution.

What I'm looking for

Here's a short, targeted answer for "What are you looking for": Short version: A Senior/Founding AI Engineer role where I can build and ship agentic AI systems — LLMs, RAG, multi-agent orchestration, and RL — in production, not just prototypes. I'm looking for a team that cares about the unglamorous parts too: safety guardrails, human-in-the-loop design, and making agentic systems reliable at scale. Open to remote roles that value depth and ownership over years-on-pape

Experience

Energymind

AI Systems Engineer (Reinforcement Learning)

Energymind

Architected a centralized AI energy-governance system combining an LLM as high-level planner (interpreting hostel needs, crisis context, and policy rules) with a DQN/RL agent executing low-level, room-level energy allocation decisions. Designed the LLM layer to add explainability, planning, and human-aware reasoning on top of a centralized DQN that delivers fast, consistent, optimized control across cost, comfort, fairness, demand satisfaction, and carbon-aware usage objectives. Built the full pipeline end-to-end: LLM planner → centralized policy layer → DQN/RL executor → smart hostel environment → metrics, grader, visualization, and feedback loop for continuous improvement. Ran a data-driven evaluation pipeline benchmarking the RL agent against random and heuristic baselines across total cost, demand satisfaction, violations, fairness, complaints, carbon usage, and trust metrics.

LLMsDQNReinforcement Learning

Education

National Institute of Technology Agartala

National Institute of Technology Agartala

B.Tech · Electronics and Communication Engineering

2024 – 2028

NIT Agartala ,Tripura, Agartala, 799046, India

B-TECH · ECE

2024 – 2028

Nav Jeevan English School, Belthra

Senior Secondary (XII) · CBSE

2023
Nav Jeevan English School

Nav Jeevan English School

Secondary (X) · CBSE

2021

Skills

raglanggraphllmplatfom engineerai safetyAi evalsmachine learningmlopsAgentic aipython