CS undergrad building production-grade systems — distributed backend infrastructure in Go to ML platforms with real deployments.
Currently: 6th semester @ PES University, Bengaluru | Expected graduation 2027
Looking for: Backend or ML engineering internships where I can contribute from day one
Gitlore — NL Question-Answering over GitHub Repos
Temporal Knowledge Graph system for grounded natural-language querying over codebases. Combines TKG traversal, BM25, and vector search with a cross-encoder reranker — no hallucinated history.
- Insight: Commits and PRs are inherently temporal; TKG + general LLM outperforms specialized models for this domain
- Tech: Python, FastAPI, Kuzu, ChromaDB, BM25, Groq, LLaMA 3.3-70B, React, Vite, Tailwind
MiniRAFT — Raft Consensus Engine
Raft consensus implementation in Go — leader election, WAL-fsynced log replication, and single-RPC catch-up across a 3-replica cluster.
- Tech: Go, gRPC, Protobuf, Docker
SensorLens — Manufacturing Anomaly Detection
Anomaly detection platform for CNC manufacturing with Isolation Forest and LLM-generated plain-English diagnostics.
- Live: sensorlens.streamlit.app
- Tech: Python, FastAPI, scikit-learn, Groq, LLaMA 3.3-70B, Streamlit
Multimodal agentic system for real-time ICU deterioration detection — ECG, hemodynamic vitals, computer vision, and clinical notes. Built a rejection sampling pipeline enforcing physiological consistency across 2,198 synthetic profiles with zero train/test leakage.
- Tech: Python, PyTorch, scikit-learn, Neo4j, FastAPI
Languages: Go · Python · Java · C/C++ · SQL · JavaScript/TypeScript
Backend: FastAPI · gRPC · Protobuf · REST API · WebSockets · Docker
Databases: PostgreSQL · Redis · SQLite · Kuzu · ChromaDB
ML/AI: scikit-learn · PyTorch · TensorFlow · pandas · NumPy · Groq · LLaMA · HuggingFace
Tools: Git · k6 · pytest · Linux · Prometheus
- LinkedIn: linkedin.com/in/shrish-singh-thakur
- Email: shrishsinghthakur04@gmail.com
- Location: Bengaluru, Karnataka, India
💡 Open to backend/ML engineering internship opportunities · Building systems that scale

