the anti job boardactive 4mo ago
MOHD KAMRAN TAHIR
Founder & Lead Backend Engineer @ LeapOCR.com
Summary
Mohd Kamran Tahir is a seasoned Backend-focused Senior Software Engineer with over 5 years of experience in architecting high-throughput distributed systems and scalable microservices. He has a proven track record in optimizing cloud infrastructure and building data pipelines capable of processing over 100,000 events per minute, showcasing his expertise in Golang, Node.js, and AWS.
Looking for
Skills
Preferences
Open to
everywhere
The story
What MOHD is looking for
My dream role is being the backend anchor in a fast-moving, high-agency team, like a founding engineer or lead builder. Day-to-day, I want to be architecting systems that actually matter. Not just building basic CRUD apps, but handling heavy concurrency, asynchronous AI agent orchestration, and distributed state. I completely lose track of time when I'm deep in a performance rabbit hole. Give me a flame graph, a slow Go or Rust microservice, and a bottleneck to parallelize, and I will happily spend hours optimizing it until the latency drops from seconds to sub-50 milliseconds. I thrive in a culture of high ownership where we debate architecture fiercely, ship incredibly fast, but deploy bulletproof, fault-tolerant code to production.
Why MOHD is exploring
I recently built my own AI SaaS (LeapOCR) from the ground up. I architected the infrastructure, wrote the asynchronous AI workflows, got it scaling to handle 50k+ documents a month, and proved I can build a complex product end-to-end. But honestly? The solo founder/indie hacker grind can be a lonely echo chamber. I miss the velocity, the friction, and the sheer momentum of a highly talented engineering team. I want to take this "zero-to-one" founder muscle I’ve developed and apply it to a massive, high-impact problem alongside a cracked team of people who are smarter than me, where we can build something monumental together.
What won't fit on a resume
My resume says "Implemented Temporal.io," but it doesn't show the massive 3-week rabbit hole I went down because standard message queues (like Redis/Nats/Kafka) kept failing me on long-running LLM inferences. I was so frustrated by dropped payloads and API timeouts that I completely tore down my architecture and taught myself Temporal to achieve deterministic, fault-tolerant state management. I also constantly build things just to scratch my own itch. Recently, I started hacking on an AI-driven developer workflow tool (I called it CallZero) purely because I was annoyed by the friction in my own SDLC. I’ve spent weekends writing custom AI agent skills and playing with MCPs (Model Context Protocol) to let LLMs talk directly to local databases, just to see how far I could push autonomous reasoning. Finally, I’m the guy who gets irrationally annoyed by a slow deployment, I will routinely stay up till 3 AM rewriting Dockerfiles and caching layers just to shave 5 minutes off a CI/CD pipeline, even if nobody asked me to.
Work & Projects
The Anti Job Board
Skip the application black hole. Get warm intros to founders hiring now.
Create your profile