Looking for a job

Aayush Agarwal

@aayushagarwal2

AI Engineer at OGGNAI TECHNOLOGIES building real-time CCTV analytics and agentic sales prospecting systems.

At OGGNAI TECHNOLOGIES, I developed a real-time CCTV analytics platform using NVIDIA DeepStream, with person tracking across video frames and camera views. I also built reliable analytics data pipelines to support footfall and entry/exit analysis.

I developed an AI sales prospecting agent with FastAPI orchestration, E2B sandbox execution, and PostgreSQL-backed approval workflows. My projects include a CLIP-based fashion image recommendation system deployed on AWS and a LoRA fine-tuned vision model; I also co-authored an IEEE ICIC3S 2026 paper on physics-informed federated learning.

Experience

Work history, roles, and key accomplishments

Current

AI/ML Intern

Current

OGGNAI TECHNOLOGIES

May 2026 - Present (6 months)

Developed a real-time CCTV analytics platform using NVIDIA DeepStream, integrating IP cameras through RTSP/ONVIF streams, and implemented person tracking and re-identification. Built an AI sales prospecting agent using the Pi Coding agent runtime, with FastAPI orchestration and PostgreSQL-backed approval workflows.

Nvidia DeepStream RTSP ONVIF People Tracking Re Identification Data Pipelines fastAPI E2B PostgreSQL AI Agents

AI Intern

Obsidian Technology

Apr 2025 - Jun 2025 (2 months)

Designed and streamlined multiple AI agents for KYC automation, RAG-based customer support, and chat-based image generation with age-specific conditioning using N8N, with seamless integration via webhooks for real-time data exchange.

AI Agents KYC Automation RAG N8n Webhooks Image Generation