Case studies
Production AI, end to end
Selected work spanning edge agentic AI, multimodal Video-RAG, inference optimization, enterprise AI workflows, and applied computer vision — each framed around the problem, what I built, and measurable impact.
Agentic AI + Edge AI + Computer Vision
Production Edge Agentic AI for Mining Safety
Designed real-time on-device agentic AI systems for safety-critical video analytics on NVIDIA Jetson AGX.
- 45% reduction in manual monitoring
- 30 FPS real-time video analytics
- Improved safety monitoring reliability
Multimodal RAG + LLM Systems
Video-RAG for Continuous Video Search
Architected retrieval-oriented video understanding pipelines for natural-language search over live and long-horizon video streams.
- Sub-second search latency over multi-day indexed video
- Contextual querying and decision support over live video streams
- Improved explainability through retrieval-backed reasoning
Agentic AI + Enterprise Workflow Automation
Enterprise AI for Sprint Planning and Resource Allocation
Developed an agentic planning workflow to support prioritization, sequencing, trade-off analysis, and resource allocation.
- Improved visibility into dependencies and constraints
- Supported AI-assisted prioritization and sequencing
- Framed as a human-AI collaboration system for workflow redesign
ML Systems + Inference Optimization
High-Performance Inference and MLOps for Real-Time AI
Built GPU-accelerated inference and deployment pipelines that improved throughput, latency, reproducibility, and validation.
- 2x throughput improvement
- 60% latency reduction
- 50% IPC latency reduction
Computer Vision + Applied ML
Real-Time Infrastructure Defect Detection
Built real-time anomaly detection and inspection systems integrating live drone feeds, ML predictions, and analytics dashboards.
- 90% accuracy
- 4x faster processing
- 30% lower inspection costs
LLM Automation + Developer Productivity
LLM-Based Pull Request Review Assistant
Built an LLM-based pull request review system to automate code analysis and improve engineering efficiency.
- Improved engineering efficiency
- Supported faster applied AI development workflows