Spark Hack Series — New York (NVIDIA)
FloodIQ — end-to-end flood planning stack (FastAPI + React + LLM advisories).
Open to Summer 2026 internships
Sudharshan Rao Allu
ML & Backend Engineer @ NYU
MS Computer Engineering · Python, ML pipelines, REST APIs
New York, NY
About
I'm Sudharshan Rao Allu, an MS Computer Engineering student at NYU Tandon. I build backend services and ML systems end-to-end—Python, REST APIs, Docker/Kubernetes, and production-style workflows from experimentation to deployment.
Highlights
- RAG + LLM systems: production-style document assistant (LangChain, ChromaDB, hybrid retrieval) and LLM fine-tuning/serving work (LoRA + 4-bit quantization).
- MLOps & serving: MLflow/DVC pipelines, FastAPI inference, Dockerized services, and CI/CD-minded iteration.
- Systems + security research: NYU Secure Systems Lab contributor on Lind (NaCl/WebAssembly syscall hardening) with benchmarking-focused validation.
- Applied ML + signals: CNN/Transformer diagnostics (MRI) and IoV intrusion detection (up to 99.9% F1 on public benchmarks); predictive maintenance modeling with time-series + FFT features.
- Infrastructure engineering: Terraform/Ansible + Kubernetes platform work for reproducible deployments and observability.
- Hackathons: shipped full-stack + AI projects across NVIDIA Spark Hack NYC, Google Build-with-AI @ NYU, and CogniChip @ NYU (details).
I'm currently seeking Summer 2026 internships in Backend or ML Engineering.
Updates
| Apr 2026 | Spark Hack Series — New York (NVIDIA): built FloodIQ (details) |
| Mar 2026 | NYC Build With AI Hackathon (Google) @ NYU Tandon: built Mirror (details) |
| Feb 2026 | CogniChip Hackathon at NYU: built ARCH-AI (details) |
| Sep 2025 | Research Assistant at NYU Secure Systems Lab (Lind, WebAssembly / RustPOSIX) |
| Sep 2025 | IT Co-Head, Hemmes Engineering Science Society (HESS) at NYU |
| Aug 2025 | Started MS in Computer Engineering at NYU Tandon |
| — | Open to Summer 2026 internships (Backend / ML Engineering) |
Hackathons
Recent hackathons and what we built. View all hackathons →
NYC Build With AI Hackathon
Mirror — AI fashion stylist Chrome extension (vision + voice + shopping search).
CogniChip Hackathon at NYU
ARCH-AI — RL + LLM guided microarchitecture optimization (Yosys synthesis metrics).
Selected projects
ML systems, RAG, and backend work. View all projects →
Real-Time Recommendation System with MLOps
Full ML pipeline with MLflow/DVC, hyperparameter tuning (R² -0.26 → 0.28), FastAPI + Docker deployment.
- MLflow
- DVC
- FastAPI
- Docker
RAG-Powered Document Assistant (Smart RAG)
Production RAG for Q&A over PDF/DOCX/TXT/MD: ChromaDB, hybrid retrieval, multi-turn chat, source citations, analytics. Streamlit + Docker, live on Streamlit Cloud.
- LangChain
- RAG
- ChromaDB
Transfer Learning and CNN-Based IDS
IoV intrusion detection with CNN ensembles, VGG/Inception/Xception transfer learning. 99.9% F1 on Car-Hacking, CICIDS2017.
- PyTorch
- Transfer Learning