Anup Hegde

Architected an Agentic AI contract drafting and editing system using LangGraph and FastAPI, enabling automated generation alongside localized, section-level user revisions. Engineered targeted context-routing for granular edits, drastically reducing Gemini API token consumption and latency by isolating relevant clauses instead of reprocessing whole documents. Managed dynamic state persistence with MongoDB and instrumented continuous workflow tracing via LangSmith to evaluate prompt performance across complex graph transitions.

AI Developer Intern at Melento (formerly SignDesk) (2026-08 – Present)

• Architected an Agentic AI contract drafting and editing system using LangGraph and FastAPI, enabling automated generation alongside localized, section-level user revisions.
• Engineered targeted context-routing for granular edits, drastically reducing Gemini API token consumption and latency by isolating relevant clauses instead of reprocessing whole documents.
• Managed dynamic state persistence with MongoDB and instrumented continuous workflow tracing via LangSmith to evaluate prompt performance across complex graph transitions.

AI/ML Intern at Tattvira Private Limited (2026-04 – 2026-07)

• Engineered a real-time computer vision pipeline for a live surveillance system, deploying continuous monitoring algorithms to automate security detection.
• Annotated and curated 1,500–2,000 images across video datasets using CVAT and Roboflow, applying advanced augmentation and preprocessing to boost model robustness.
• Trained and fine-tuned edge architectures (RTMPose, YOLOX) and converted models to ONNX for optimized deployment, balancing low-latency inference speeds with high accuracy.

B.E in Computer Science (Data Science) in Computer Science – Sai Vidya Institute of Technology (2023 – 2027)

12th Standard (PCMB) in PCMB – MES PU College (2020 – 2022)