π mandeep0402.github.io Β· πΌ LinkedIn Β· π« mandeepjaglan04@gmail.com
AI & ML engineer with 4+ years of experience across data analytics, NLP pipelines, computer vision systems, and MLOps infrastructure. Currently an AI Infrastructure Engineer at Streamingo.ai, where I build end-to-end video processing pipelines, RAG-based SQL intent systems, OCR/ReID workflows, and scalable ETL infrastructure on GCP. B.Tech in Computer Science from IIIT Bhagalpur.
- Video Processing & ML Infra β RabbitMQ + Redis pipelines orchestrating VLM inference on 2,000β4,000 videos/day across multi-GPU fleets; Jenkins CI/CD from video recording to results delivery
- Computer Vision β basketball broadcast β top-down 2D minimap pipeline with YOLOv9/v11, SAM 2, BoxMOT, homography, Kalman filtering, team ReID & jersey OCR
- NLP & LLM β natural language β intent β BigQuery SQL pipelines with schema-aware prompt engineering and automated chart/insight generation
| Package | Description |
|---|---|
| VideoProcToolkit | Video download, multi-backend frame extraction (FFmpeg/OpenCV/Decord), trimming, concatenation, FPS/resolution conversion, thumbnails & metadata |
| basketball_datakit | NBA/NCAA play-by-play, box scores, event classification, lineup tracking, ESPN scraper fallback & name normalization |
- Languages: Python, SQL, PySpark, C/C++, Shell Scripting
- Data & Cloud: BigQuery, GCP/GCS, AWS S3, Snowflake, Hadoop, Hive, Docker
- MLOps & Infra: Jenkins CI/CD, RabbitMQ, Redis, ETL Pipelines
- CV & ML: YOLOv9/v11, SAM 2, BoxMOT, Homography, OCR, ReID, Kalman Filter
- NLP & LLM: RAG Pipelines, Prompt Engineering, LLM Fine-tuning, Claude API
- Libraries & Tools: Pandas, NumPy, Scikit-learn, TensorFlow, FFmpeg, OpenCV, Git, Linux, Power BI
- AI Infrastructure Engineer β Streamingo.ai (Oct 2024 β Present)
- SDE 2, Machine Learning β Nanonets, via Crewscale (Mar 2024 β Sep 2024)
- Associate Data Analyst β Comscore (Jan 2022 β Feb 2024)