Aditya Raut
GenAI Engineer shipping production RAG pipelines and edge ML
New to Contra
Aditya is ready for their next project!
Title: Production ML Inference Service — Sales Forecasting
Description: XGBoost model (R²=0.9247) with ONNX Runtime inference. Reduced user inputs 84% via auto-enrichment. Deployed with Flask + Docker + Kafka streaming on Render with real-time observability.
0
5
3. Haystack OSS PR #11419
Title: Open Source Contribution — deepset-ai/Haystack
Description: Fixed DocumentLanguageClassifier crash on blob-only documents. Merged into a 25k+ star production framework used in enterprise RAG pipelines globally.
0
9
2. Wheat Disease Intelligence Platform
Title: Edge AI Deployment — Wheat Disease Detection
Description: ConvNeXt-Tiny (88.46% accuracy), INT8 ONNX quantization cutting model size 75%, 65ms CPU inference. Full stack with human-in-the-loop feedback pipeline.
0
14
1. RAG Modernization at AllCognix
Title: Production RAG Pipeline Optimization
Description: Migrated legacy RAG system to Haystack 2.x + GPT-4o. Cut latency 40%, token costs 60%, ingestion from 70s to 27s in live production.
https://rag.allcognix.com/
0
16