Skip to content

Open to full-time roles

Bengaluru, India IST (UTC+5:30)

Venkat Swaroop, full-stack and AI engineer

Full-stack & AI engineer who ships real products.

  • AI agents
  • MCP servers
  • LLM APIs
  • DevOps

I’m the founder of NVS Ventures. I design, build and run live products end to end: AI agents and MCP servers, PostgreSQL schemas and mail servers, CI/CD and production deploys.

View work Resume GitHub (opens in a new tab)LinkedIn (opens in a new tab)

Live now

Proof of work

Shipped withGKVK · UAS BangaloreXreacherVahanBay
automated DMs a day across 50+ accountsXreacher
10K+
uptime for microservices on Cloud RunGKVK
99.5%
release cycle after the CI/CD rebuildGKVK
3d → <4h
engineers led on geospatial dashboardsGKVK
4
downloads of the VahanBay Android appGoogle Play
10K+
  • Next.js
  • React
  • TypeScript
  • Python
  • FastAPI
  • Django
  • Node.js
  • PostgreSQL
  • MongoDB
  • Redis
  • Celery
  • Docker
  • Nginx
  • GitHub Actions
  • GCP Cloud Run
  • Prometheus
  • Grafana
  • Claude
  • MCP
  • LangGraph
  • Gemini API
  • Cloudflare

Selected work

Products I designed, built and still run.

Live products people use today: two SaaS products I run under NVS Ventures and VahanBay, a vehicle marketplace on the web and Google Play. Plus research and side projects.

Agent-native email infrastructure

Live
Live site

Koltrix

A team inbox on your own domain, mailboxes for AI agents, an email API and an MCP server, in one product I designed, built and run.

  • Multi-domain inbox
  • Transactional API
  • SMTP relay
  • Newsletters
  • MCP server for Claude, ChatGPT, Cursor and Claude Code over OAuth
  • Each AI agent gets its own address and a scoped key; agents draft, people send
  • Per-organization PostgreSQL schemas, Postfix, DKIM and Rspamd, HMAC-signed webhooks

Voice-calling reminders

Live
Live site

ReminderIt

Text it on WhatsApp what to remember; at the right local time your phone rings, in one of 14 languages, with a WhatsApp backup.

  1. Schedule
  2. Voice call
  3. WhatsApp fallback
  • MCP server, so Claude and ChatGPT can create and manage reminders
  • Timezone-aware call scheduling with Celery and Redis
  • Automatic WhatsApp fallback with Done and Snooze

Vehicle marketplace · Web + Android

Private codeLive
Live site

VahanBay

A marketplace to buy and sell trucks, buses, cars, bikes, tractors and spare parts across India, with a directory of dealers, garages, transport services and RTO agents, on the web and as an Android app.

10K+downloads on Google Play

  • Vehicle listings
  • Business directory
  • Chat & enquiries
  • Jobs
  • Android app
  • Listings with filters for category, location, price and year; buyers enquire and chat with sellers
  • Seller verification with KYC, phone OTP sign-in, favourites and notifications
  • Finance, RTO services and automotive jobs in the same app

Patent application published

AI agricultural advisor robot · GKVK, UAS Bangalore (COE-WM)

LRI Mitra

A multilingual, voice-interactive farming robot that reads soil report (LRI) cards with a vision-language model and recommends crop-specific fertilizer doses.

ಕನ್ನಡ · English · हिन्दी · తెలుగు

  1. Read LRI card
  2. Vision-language model
  3. Fertilizer advice
  • Vision-language model
  • LLM agent
  • Speech
  • Ubuntu
  • Hardware
Patent details (LRI Mitra)
More on GitHub (opens in a new tab)

AI engineering

AI that works on real systems, with the guardrails to ship it.

Agents, MCP servers and LLM features running inside my own products, not just demos.

Agents that do real work

In LRI Mitra, an agentic engine reads a farmer’s soil-report card through a camera with a vision-language model, recommends crop-specific fertilizer doses and answers by voice in four languages. In Koltrix, agents sort the inbox, screen cold pitches and draft replies.

  • Vision-language model reading real documents (LRI cards)
  • Multilingual voice in Kannada, English, Hindi and Telugu
  • Inbox agents that sort, screen and draft

Shipped in

LRI Mitra · Koltrix

  • Claude API
  • Gemini API
  • LangGraph
  • Python

Try it: an assistant using my MCP servers

Choose a prompt and watch the tool calls, including the ones my guardrails refuse.

Pick a prompt

Example trace, not a live call. The tool names are the real ones my MCP servers expose; the data is made up.

assistant · ReminderIt MCP
  1. userRemind me to call Mom at 8 PM every day
  2. thinkA daily phone reminder at 20:00, in the user’s own timezone.
  3. tool
    whoami()
    {}

    { "timezone": "Asia/Kolkata" }

  4. tool
    create_reminder()
    {
  "title": "Call Mom",
  "cron_expr": "0 20 * * *",
  "timezone": "Asia/Kolkata"
}

    { "status": "scheduled", "next_call": "today, 8:00 PM IST" }

  5. replyDone. Your phone will ring every day at 8:00 PM IST, with a WhatsApp backup if you miss it.

How I ship

From an idea to something people use, and keeping it up.

The same loop for my own products and for teams I’ve worked with.

  1. 01

    Design

    Start from the user and the data: flows, schemas and API contracts before code.

    Per-org PostgreSQL schemas in Koltrix

    • PostgreSQL
    • MongoDB
  2. 02

    Build

    Typed frontends and Python backends, with tests where they protect real behaviour.

    Next.js, FastAPI and Django in production

    • Next.js
    • FastAPI
    • Django
  3. 03

    Deploy

    Containers and CI/CD, so shipping is boring and frequent.

    Release cycle cut from 3 days to under 4 hours

    • Docker
    • GitHub Actions
    • GCP Cloud Run
  4. 04

    Operate

    Metrics, alerts and error tracking, and I’m the one who fixes it.

    99.5% uptime on Cloud Run

    • Prometheus
    • Grafana
    • Sentry

Experience

From intern to founder, shipping to production at every stop.

Startups, a university research centre and my own company. Small teams, real users, and a lot of ownership.

  1. May 2026 – Present

    Bengaluru

    Now

    NVS Ventures

    Founder & Full Stack Engineer

    • Designed, built, launched and operate Koltrix and ReminderIt as the sole engineer: architecture, backend, frontend, deployment and billing.
    • Run production on an Ubuntu VPS with Docker Compose, Nginx, Redis and MongoDB, tuning memory limits and swap to keep several services stable on one host.
    • Handled the business side end to end: MSME (Udyam) registration, Razorpay onboarding and domain strategy.

  2. Jun 2025 – Oct 2026

    Remote

    Xreacher

    Full Stack Developer

    automated DMs / day
    10K+
    accounts orchestrated
    50+
    • Architected a chatbot-based messaging system in Next.js, Python and MongoDB with zero message drops.
    • Built a lead-scraping pipeline and a multi-account orchestration engine for the sales team.
    • Set up Prometheus and Grafana monitoring for the PM2-managed process fleet.

  3. Aug 2024 – Jun 2025

    Hybrid · Part-time

    GKVK Watershed Management

    Lead Full Stack & DevOps Developer

    uptime on Cloud Run
    99.5%
    release cycle
    3d → <4h
    engineers led
    4
    • Led a team of 4 building real-time geospatial dashboards for watershed field data (React, Firestore).
    • Designed and deployed containerized microservices on Google Cloud Run.
    • Built a CI/CD pipeline with automated tests and a staging environment.

  4. May 2024 – Aug 2024

    Remote

    Bryckel.AI

    Full Stack Intern

    • Shipped end-to-end features across a React frontend and Django backend, including the user onboarding flow and an analytics dashboard.
    • Wrote unit and integration tests with Jest and Pytest.

  5. Jan 2024 – May 2024

    Hybrid

    Skillmithra.com

    Lead Full Stack Developer · Freelance

    • Independently architected and deployed a learning platform on Django, PostgreSQL and AWS EC2.
    • Built an admin dashboard with role-based access control and real-time analytics, served via Gunicorn and Nginx with SSL.

Patent & achievements

Recognised for building things that work.

A published patent application from my work with GKVK, and national-level hackathon wins.

Patent application published

Indian Patent Application No. 202641127012 A

AI-Based Multilingual Voice-Interactive Farming Robot and Method for Personalized LRI Card-Crop-Fertilizer Advisory

Role
Co-inventor
Filed
25 Oct 2026
Published
30 Oct 2026
Status
Pending examination

Joint application by the Centre of Excellence on Watershed Management, UAS (GKVK), and Dayananda Sagar College of Engineering.

Skills

The tools I use to take a product from schema to production.

Frontend

  • React
  • Next.js
  • TypeScript
  • JavaScript
  • Tailwind CSS
  • Chart.js
  • WebSockets

Backend

  • Python
  • Django
  • FastAPI
  • Node.js
  • Flask
  • REST APIs
  • Celery
  • Django Channels

Databases

  • PostgreSQL
  • MongoDB
  • MySQL
  • Redis
  • Firestore
  • DynamoDB

Cloud / DevOps

  • Docker
  • Docker Compose
  • GitHub Actions
  • GCP Cloud Run
  • AWS (EC2, S3)
  • Nginx
  • Gunicorn
  • PM2
  • Linux
  • Prometheus
  • Grafana
  • Sentry

AI / Integrations

  • Claude API
  • Gemini API
  • MCP
  • LangGraph
  • Whisper
  • Twilio
  • Razorpay
  • WhatsApp

Hardware

  • ESP32
  • Arduino
  • IoT sensors
Portrait of Venkat Swaroop in a dark shirt against a black background

Bengaluru · IST

About

I build and run my own products from database to deployment. I learn fast, take ownership, and finish what I start.

I graduated in Computer Science from Dayananda Sagar College of Engineering in 2026. Today I run NVS Ventures, where Koltrix and ReminderIt serve real users on infrastructure I set up and look after myself.

Before that I led a small team building watershed dashboards at GKVK and worked on outreach automation at Xreacher. Next, I want to join a product team where I can own features end to end: backend, frontend and the pipeline in between.

Based in
Bengaluru, India
Education
B.Tech CSE, DSCE (2022–2026)
Company
NVS Ventures (Udyam-registered)
Looking for
Full-stack, backend or AI-application engineering roles

FAQ

Questions recruiters usually ask.

What roles are you looking for?

Full-time full-stack, backend or AI-application engineering roles at startups and product companies, where I can own features end to end.

Where are you based?

Bengaluru, India. My working hours are in IST (UTC+5:30).

Have you run software in production?

Yes. Koltrix and ReminderIt are live products I built and operate myself on Ubuntu with Docker Compose, Nginx, Redis and MongoDB. I also built VahanBay, a vehicle marketplace on the web and Android with 10K+ downloads on Google Play. At GKVK I ran microservices on Google Cloud Run at 99.5% uptime, and at Xreacher the messaging system I architected handled 10K+ automated DMs a day across 50+ accounts.

What AI work have you shipped?

MCP servers for both of my products, so Claude, ChatGPT, Cursor and Claude Code can work with them; agent mailboxes in Koltrix with scoped keys, human-in-the-loop sending and audit logs; and a vision-language-model agent in LRI Mitra, which led to a published patent application.

Have you led a team?

Yes. At GKVK Watershed Management I led a team of 4 building real-time geospatial dashboards, and built the CI/CD pipeline that cut the release cycle from 3 days to under 4 hours.

What is your education?

B.Tech in Computer Science and Engineering from Dayananda Sagar College of Engineering, Bengaluru (2022 to 2026).

How do I get in touch?

Email venkatnvs2005@gmail.com, or download my resume as a PDF. My code is on GitHub and I am on LinkedIn.

Contact

Hiring for full-stack, backend or AI-application roles? Let’s talk.

GitHub (opens in a new tab) LinkedIn (opens in a new tab)Download resume

Open to full-time roles. It’s IST (UTC+5:30) in Bengaluru right now.

esc
↑↓ to move · ↵ to runvenkatswaroop.com