Open to opportunities

Marda Anurag

@mardaanurag

Message

I build production-ready AI systems, from fraud detection research to multi-agent applications.

India

Message

What I'm looking for

I'm seeking AI/ML Engineer and Software Development Engineer roles that combine research depth with production-grade implementation, especially opportunities to build and deploy machine learning systems and LLM-based agentic applications.

I built GRAFT, a faculty-reviewed game-theoretic reinforcement learning approach for fraud detection on the Bitcoin OTC trust network. It achieved 99.6% fraud-detection accuracy, 100% recall, and increased routing success from 50% to 100% at deployment.

I'm completing a B.Tech in Computer Science and Engineering at VNIT Nagpur while building ML architectures from scratch, including Q-learning policies, Neuro-Fuzzy inference systems, and ensemble classifiers.

My projects span an agricultural multi-agent system using Google ADK, Gemini, FastMCP, real-time weather and market data, and a multilingual bus announcement system with PyTorch-based ETA prediction. I've also delivered end-to-end Flask, React.js, REST API, and Oracle SQL applications.

I enjoy combining research depth with production-grade implementation, from low-level C systems such as heap allocators and arbitrary-precision arithmetic libraries to LLM agent pipelines using LangChain.

Experience

Work history, roles, and key accomplishments

Junior Scientist Organizer

AXIS Fest, VNIT Nagpur

Aug 2024 - Oct 2024 (2 months)

Spearheaded the Junior Scientist track at AXIS Fest, coordinating 5-10 sub-events with ~500 participants from planning through execution.

Event Planning Coordination Leadership

Social and Health Secretary

Boys Hostel, VNIT Nagpur

Jul 2023 - Nov 2023 (4 months)

Served as welfare representative for ~400 hostel residents, resolving health, social, and administrative concerns in coordination with institute authorities.

Problem Solving

Machine Learning Engineer

VNIT Nagpur

Final-year B.Tech student with research in game-theoretic reinforcement learning for fraud detection, achieving 99.6% accuracy. Experienced in building custom ML architectures and deploying them end-to-end.