Sneha Gawali

000

SNEHA GAWALI

AI · Robotics · Engineering

Robotics and AI engineer at VTU. I build systems that sense the room, make a decision, and move with intent. Graph optimization, deep RL, and offline RAG, shipped with real constraints.

Projects

IncidentRAG

on-call triage

IncidentRAG

Offline RAG agent: Oracle 26ai HNSW search plus a ReAct tool loop on Llama 3.1 8B. Sub-second retrieval, no cloud API cost.

Python / LangChain / Llama 3.1 / FastAPI / Docker

DisputeFlow

AI / application

DisputeFlow

Operations console for payment-dispute analysis. Six LangGraph nodes classify a case, score supplied facts, draft a rebuttal, and record a simulated filing. Win probability is XGBoost on synthetic labels.

Python / LangGraph / XGBoost / React / PostgreSQL

RebalanceRL

deep RL

RebalanceRL

PPO agent in a custom bike-sharing sim with 300+ stations. Trip success 80.60% against a 5.56% non-learning baseline.

Python / PPO / Stable-Baselines3 / Flask / React

DeepShield AI

deepfake / live video

DeepShield AI

Real-time deepfake detection for live video calls. Behavioral and voice signals feed a live risk score, confidence metrics, and visual alerts, with OBS Virtual Camera support.

TypeScript / Vite / MediaPipe / RNNoise

Experience

  • 2026

    GRAG AI Factory (startup)

    AI Research Analyst Intern

    Clinical ML pipeline on de-identified screening records. Gradient-boosted and deep classifiers, FastAPI/Docker inference.

  • 2026

    Rooman Technology

    Machine Learning Intern

    End-to-end ML pipelines and REST APIs on AWS. Regression, classification, clustering for technical and business teams.

B.E. Robotics and Artificial Intelligence. Maratha Mandal Engineering College, VTU. CGPA 8.2 / 10 · 2022-2026.

Let's connect.

Copiedsneha.gawali19@gmail.com

GitHubLinkedInResume