Manish

Manish Kumar Sharma

Applied AI, backend systems, and the web apps around them.

I'm an applied AI engineer at InvestConclave and a final-year computer science student at the National Forensic Sciences University, specialising in cyber security.

I like systems that have to work under load: a video pipeline that turns a ten-minute job into ninety seconds, a retrieval system over forty thousand court judgments, a deployment platform with sandboxed build workers. Lately most of that energy goes into AI agents and the plumbing that keeps them honest.

Since 2024 I've also shipped a dozen client sites as a freelancer, from booking platforms to e-commerce, and kept them running at 99.9% uptime.

What I do

AI agents
Agents and LLM features that do real work inside products: retrieval, tool use, evaluation, and the guardrails that make them safe to ship.
Backend systems
Go, Node.js and Python services. Queues, caching, and processing pipelines for video and documents, built to be measured and scaled.
Full stack web apps
React and Next.js on top of PostgreSQL and Prisma. Dashboards, auth, payments, scheduling, and the deploys that keep them up.
System design
Thinking through failure modes, idempotency and data flow before they turn into incidents.

Experience

  1. 2026 – Present

    Applied AI Engineer, InvestConclave

    • Building applied AI features and agent workflows for the platform.
  2. Dec 2025 – Jan 2026

    Engineering Intern, Nowlez (internship)

    • Designed the backend architecture and document ingestion pipelines for legal AI analysis, including OCR for uploaded documents.
    • Built secure API endpoints with data protection practices aligned to legal-tech compliance, and shipped two web apps in the process.
  3. Aug 2025 – Jan 2026

    Full-Stack Developer, Scalez Media (contract)

    • Led backend architecture at an early-stage startup, building the production system from the ground up.
    • Built a high-throughput video processing microservice in Python and FFmpeg that cut turnaround per video from 10 minutes to 90 seconds.
    • Built React dashboards for monitoring jobs and scheduled publishing, and owned deployment and scaling on DigitalOcean.
  4. Jan 2024 – Present

    Freelance Developer, Self-employed

    • Delivered 12 end-to-end client projects, from booking platforms to e-commerce, with React frontends and Node.js backends.
    • Designed relational schemas and API layers, integrated payments and notifications, and kept production apps at 99.9% uptime.

Projects

  • The Cyber Bird

    My own business. Web development, SEO and applied AI for small companies.

  • Legal RAG for Indian judgments

    Retrieval system over 40,000+ Indian court judgments for accurate legal citation. Python ingestion, chunking and embeddings; Go services with Kafka for distributed ingestion and Redis for hot retrieval paths; a React dashboard for citation search and result inspection.

  • Vercel-style deployment platform

    An automated CI/CD platform with S3 artifact storage, Redis-backed job queues and sandboxed parallel build workers. Node.js on AWS.

Client work

Twelve client sites since 2024, from law chambers to ticket booking, with SEO handled end to end.

Education

2022 – 2027

B.Tech + M.Tech in Computer Science, specialization in Cyber Security

National Forensic Sciences University. Integrated dual degree, expected 2027.

Writing

All posts

Skills

Languages
Go, TypeScript, JavaScript, Python, C++, SQL
Backend
Node.js, Hono, Express, REST APIs, Microservices, Kafka, FFmpeg
Frontend
React, Next.js, Tailwind CSS
Databases
PostgreSQL, MySQL, MongoDB, Redis, Prisma
Infrastructure
Docker, AWS (EC2, S3, IAM, Auto Scaling), DigitalOcean, CI/CD
Focus areas
System design, API design, Auth (JWT, OAuth), Performance, Applied AI