Engineer & Builder · Dublin, Ireland

Vatsal
builds
things.

Analytics and ML at Google by day. Building things in the evenings. Ten years across data engineering, ML, and production systems. Two live apps and a homelab that runs itself.

10yr In Tech
Google T&S Ads
3 Apps Live
Vatsal Mishra
Vatsal Mishra
Role Senior Engineering Analyst, Google
Background Analytics · ML · Data Eng · 10+ years
Based in Dublin, Ireland
In this issue
Selected Work 2023 – 2026
01
Consumer · AI Detection · Mobile

FakeOut: AI Image & Misinformation Detector

Fact-checking and AI-generated image detection in a single tap. Built solo, shipped on Android first, then iOS. Reached 14K+ downloads organically (no ad spend, no influencer coverage) through word of mouth and YouTube. Backend connects multiple LLM APIs into a lightweight, fast verification pipeline.

14K+ Downloads
4.2★ Play Store
iOS Now Live
Visit fakeout.io Play Store App Store
FAKEOUT.IO FakeOut — AI Image Detector & Fact Checker · Detect Deepfakes Free Detect AI-generated images, deepfakes, and verify viral claims instantly. Free Android & iOS app powered by Gemini. Trusted by 14,000+ users. Check before you share.
02
Product · Generative AI · Web

Redo AI: Space Transformation

Upload a photo of any room and get AI-generated design alternatives: different styles, furniture configurations, colour palettes. Early stage, exploring consumer generative AI in interior design. Built to see how quickly a solo dev can take an LLM-powered idea to production.

Visit re-do.ai
RE-DO.AI Redo AI — Transform Any Space with AI AI-powered visualization for urban and interior design. Transform cities and homes in seconds. Try it free!
03
Product · Social · AI

Stashly: A Smarter Way to Save Reels

Built in 24 hours during a buildathon, Stashly turns Instagram Reel links into searchable cards with categories, notes, tags, and AI summaries. A small fix for the chaos of saved content.

Visit Stashly
Stashly OpenGraph preview STASHLY.REPLIT.APP Stashly — A Smarter Way to Save Reels Built in 24 hours during a buildathon. Turn Instagram Reel links into searchable cards with categories, notes, tags, and AI summaries.
Infrastructure & Homelab Ongoing
04

Home VPN

WireGuard on Raspberry Pi with Pi-hole for network-wide DNS blocking. Secure remote access to the home network from anywhere in the world.

05

NAS + Nextcloud

Self-hosted cloud storage on Ubuntu Server via Docker. Multi-device sync, backups, and sharing. No cloud subscription or vendor lock-in.

06

Home Media Server

Plex on a mini-PC with hardware transcoding. Centralised media library for all devices on the home network, no buffering.

07

Homelab Monitoring

Webmin for system administration: services, packages, users, and backups from a single web UI. Observable infrastructure without SSH depth.

08

Local AI Experiments

Ollama + OpenWebUI on a 4070 Ti Super rig. Running Llama, Mistral, and multimodal models locally for privacy-first AI experimentation.

09

Tech Content

YouTube + Instagram @techwithvatsal. Short-form tech explainers in Hinglish for the Indian tech-curious audience. AI, homelab, and data tools, explained plainly.

About
The person behind the work

Builder. Analyst. Infrastructure nerd.

"I build things to learn. Not everything ships, but the work is real."

At Google Dublin, I work on production ML and AI systems for Trust & Safety in Ads. That covers feature pipelines, data engineering, LLM governance, and post-launch monitoring at scale.

The side-project stack runs just as deep. I've been running a homelab for years: Raspberry Pis, Ubuntu mini-PCs, Docker Compose stacks. Controlling your own infrastructure teaches you things that managed cloud services actively obscure.

FakeOut came out of genuine frustration with how hard it is to quickly fact-check a screenshot or determine if an image is AI-generated. Built it in spare evenings, shipped Android first, crossed 14K downloads without any ad spend. Now live on iOS too.

Find me
GitHub LinkedIn X / Twitter YouTube Instagram
Contact
Get in touch

Let's work on something.

Working on something interesting? Questions about FakeOut? Want to talk AI, homelab, or data? Drop a message. I read everything.