Building production ML systems that actually ship — from edge devices to cloud APIs.
GenAI Engineer Intern @ AllCognix AI, building RAG pipelines and production ML infrastructure on Haystack 2.x.
- 3 merged PRs in deepset-ai/haystack (25k★) — bugs caught through production use, not code review
- Cut RAG latency by 40% and token costs by 60% via Haystack 2.x migration + context windowing
- Reduced document ingestion from 70s → 27s with parallel processing
- Published quantization stability research on Jetson Nano edge hardware — preprint on Zenodo
ML / Deep Learning
MLOps & Deployment
Backend & Data
| PR | Fix |
|---|---|
| #11419 | DocumentLanguageClassifier crash on blob-only docs (content=None) — uncaught TypeError replaced with graceful unmatched fallback |
| #11711 | RecursiveDocumentSplitter silent metadata corruption — split_idx_start miscalculated when split_unit="word"/"token" with overlap enabled; replaced unit-count arithmetic with actual overlap string character length |
| #11768 | RecursiveDocumentSplitter silent overlap loss — split_overlap ignored on no-separator fallback path in _chunk_text() |
|
Observability and diagnostics engine for Haystack 2.x RAG pipelines. Exposes document-store validation, pipeline inspection, retrieval-failure analysis, and structured debug bundle diffing via MCP.
6-class retrieval-failure taxonomy · 823-chunk live corpus · ~0.95s for 15 concurrent MCP requests |
Production ML platform with ConvNeXt-Tiny inference, CLIP-based input validation, OpenCV symptom overlays, GPT-4o mini recommendations, and a human-in-the-loop feedback pipeline.
88.46% accuracy · 75% model compression · 65ms CPU inference |
|
Quantization stability study across MobileNetV3-L, ResNet50, ConvNeXt-Tiny on Jetson Nano. Audited 14,154-image benchmark (11.6% cross-split leakage), recovered INT8 accuracy from 31.0% → 82.5%, proposed DES metric.
54.5 FPS edge inference · entropy-calibrated TensorRT · leakage-audited benchmark |
Geospatial ML on 55M rows with Haversine distance features. Containerized and served via FastAPI.
55M rows · containerized API |
- Interning @ AllCognix AI — production RAG system on Haystack 2.x, EC2, Weaviate, Redis, Vault
- OSS contributions to deepset-ai/haystack — 3 merged PRs, ongoing
- Research paper targeting Computers and Electronics in Agriculture (Q1 Elsevier)
- Open to ML Engineer / GenAI roles — remote, India & international
Pune, India · B.Tech CSE (AI & Analytics) · MIT ADT University · 2028