Deepansh Gupta

AI Engineer with a proven track record of architecting and deploying multi-agent pipelines, multi-modal conversational solutions, and edge inference applications. Expert in engineering production-grade GraphRAG systems over Neo4j knowledge bases, building autonomous closed-loop analytics engines, and executing strict data-grounding verification using LangSmith tracing. Combines deep technical prototyping and cross-attention transformer modeling with a strong background in real-time embedded AI planning to accelerate and scale enterprise-grade AI infrastructure.

End-to-End Autonomous AI Driving & Edge Inference System (2025-06 – 2026-04)

• Engineered a vision-based autonomous pipeline in CARLA, establishing high-frequency telemetry streams to fine-tune a deep CNN regression model via transfer learning for real-time steering and velocity prediction.
• Deployed real-time neural and probabilistic planning policies on NVIDIA Jetson Orin Nano using Python; benchmarked stochastic AI control (MPPI) against geometric baselines across 40 trials (~0.026m error) to co-author an IEEE manuscript.

B.Tech in Automobile Engineering – Zakir Husain College of Engineering & Technology (ZHCET), Aligarh Muslim University (2022 – 2026)