Mohd Abdul Sohail

Mohd Abdul Sohail

Tech Lead / Full Stack Engineer at Veloqo

Hyderabad, India


About

I’m a Full Stack Engineer and AI enthusiast who loves building real-world solutions with computer vision, real-time video analytics, and scalable systems. From GPU-powered inference pipelines and backend services in Python and Go to modern web apps with Angular and React, I enjoy turning complex ideas into products that work in the real world. At Veloqo, I lead a team of engineers building AI solutions for retail, smart parking, digital signage, logistics, and industrial automation across India and the Middle East. I work across the entire product lifecycle—from architecture and development to cloud deployments, edge AI, and production rollouts. I’m passionate about exploring new technologies, solving challenging engineering problems, and building things that make an impact. Always excited to connect with fellow developers, exchange ideas, and collaborate on projects that push boundaries! 🚀

Skills
Python
Go
SQL
TypeScript
JavaScript
FastAPI
REST APIs
PostgreSQL
Angular
React
GStreamer
WebRTC
MySQL
Apache Kafka
Docker
AWS
Google Cloud
Git
Kubernetes
Experience

Technical Lead – Full Stack Engineer

Veloqo IoT Ltd

Nov 2025 to Present
  • • Lead a team of engineers across backend, frontend, and AI/ML, owning system architecture, technical decisions, code reviews, and delivery from design through production deployment.
  • • Deliver end-to-end AI video analytics products spanning full stack development (FastAPI, Go, Angular, React), GPU inference, and cloud infrastructure on AWS.

Software Engineer

ZestIOT Technologies Pvt. Ltd

Nov 2022 to Nov 2025
  • • Designed and maintained REST APIs and microservices (Python, Flask, SQLAlchemy, MySQL) powering industrial computer vision products for 10+ enterprise clients.
  • • Scaled backend workloads with Celery/RabbitMQ asynchronous processing and Redis caching, reducing API response times by 90%; deployed services with Docker and Nginx on AWS and GCP.
  • • Led solution design and code reviews, and owned production debugging and customer-facing technical communication.

Python Intern

ZestIOT Technologies Pvt. Ltd

Sept 2022 to Oct 2022
  • • Developed Python REST APIs and server-side automation scripts for edge device monitoring, system health checks, and operational workflows.
Projects

Omniview – AI Video Analytics & Surveillance Platform

• Architected a distributed control-plane/data-plane platform (FastAPI, Angular, PostgreSQL, Redis, Apache Kafka) with Go services for rule evaluation, alerting, evidence capture, and notifications, supporting 500+ cameras across multiple retail stores and industrial sites in India. • Designed a decoupled ingestion and inference pipeline: GStreamer services on edge devices and CPU instances decode video and stream frames via Kafka/Redis to NVIDIA Triton (TensorRT, ONNX, YOLO, ByteTrack) on a single NVIDIA L4 GPU, sustaining 300+ FPS aggregate throughput with per-camera rates of 0.5–5 FPS tuned to each use case.

MLOps Platform – End-to-End ML Lifecycle

• Built an MLOps platform (React, FastAPI, PostgreSQL, MinIO/S3, CVAT, Celery) covering datasets, annotation, training, experiment tracking, model registry, deployment, and inference testing. • Orchestrated GPU training and serving with PyTorch, Argo Workflows, Kubernetes, KServe, MLflow, and lakeFS, with data-to-deployment lineage, project-scoped RBAC, audit logging, and Prometheus/Grafana/Loki observability.

Parking Management & Digital Signage Monitoring

• Built an edge-to-cloud platform on NVIDIA Jetson (YOLOv7 FP16, TensorRT, RTSP) with Flask, Angular 19, Node.js Socket.IO, MySQL, and Redis, monitoring parking spaces and signage displays for an enterprise client site in Dubai, UAE. • Implemented occupancy state machines, ROI management, signage OCR (pytesseract), license-plate recognition, and Twilio voice/WhatsApp alerts on edge services.

Auto Cap Short – Logistics & AI Analytics Platform

• Built backend APIs for truck analytics, cylinder defect detection, missing cap/seal detection, and automated reporting across 8 edge locations, with event-driven camera ingestion, rule evaluation, and cloud synchronization.

Steel Melting Shop Digitization – Industrial Analytics

• Reconstructed end-to-end ladle heat cycles by combining real-time PLC data with computer vision, enabling cycle-time analytics across 10 stages; supported customer validation and project sign-off.

Pure For Sure – QR-Based Traceability System

• Owned delivery from requirements to production of a cylinder traceability system with truck entry/exit analytics and truck-to-load mapping; built multi-camera QR tracking with multiprocessing for high-speed conveyors, validated on site with 1,000+ cylinders.

Education

Bachelor of Technology (B.Tech) - IT

Shadan College of Engineering and Technology

Aug 2016 to Oct 2020