Akdev Saha
hello from here
Software developer and product builder focused on shipping scalable, user-centered products by bridging engineering, product judgment, and real business constraints.
Day to day, I'm deep in the details from writing code, learning what the problem needs, and shipping reliable software to production. Sometimes I overthink the details, that's half the fun.
tip: the lofi button is real. press play if you like company.
At Face IQ Labs, I am building Clipify, the platform we use to run creator marketing and campaign operations. I work across product and engineering, shipping the systems that handle how creators submit work, get verified, and get paid.
Key focus areas include:
- Architecting and building core creator and admin systems across submissions, cashouts, verification, and campaign operations.
- Developing Meleys, an AI creator assistant with persistent chat, intent handling, rate limiting, and clear escalation into human support.
- Designing trust and safety flows for social account verification, evidence review, and risk-aware payout decisions.
- Collaborating within a lean team to build admin tooling for cashout review, creator search, video evidence, and day-to-day campaign operations.
Transitioning into this role has been an incredibly rewarding next step following my previous product-building ventures.
API observability platform for network tracing and load testing.
- Built a Go probing engine capturing DNS, TCP, TLS, and TTFB timings at sub-millisecond precision.
- Designed concurrent burst testing for 100+ parallel requests with safe resource throttling.
- Shipped a monorepo stack with Next.js, Express, Redis/BullMQ, and Go services.
CLI to scaffold containerized React + Go apps, used by 400+ developers.
- Reduced full-stack setup from hours to seconds with a single npx command.
- Pre-configured Docker, CORS, and dual hot-reload for React and Go.
- Adopted by 400+ developers for production-ready project scaffolding.
AI-powered pull request reviews for faster code quality feedback.
- Automated PR analysis with LLMs to flag bugs, quality issues, and architecture risks.
- Integrated into the review workflow so feedback lands before human review starts.
- Built as a cloud-native app with CI/CD, containerized services, and Git integrations.
I have been building and shipping products for a while, and each one has taught me something new about systems, performance, and what actually matters in real-world software.
It started with Trackly, where I focused on reliability. I built an infrastructure monitoring platform with concurrent health checks, latency tracking, and an automated alerting system. This was where I learned how important it is to design systems that do not just work, but stay reliable under real conditions.
Then I built Prwise, exploring how AI can improve developer workflows. I worked on analyzing pull requests automatically and surfacing actionable feedback before human review, which taught me a lot about developer tooling and building systems that integrate into existing workflows.
Wanting to go deeper into real-time systems, I built Flow, a communication platform supporting messaging, audio calls, and video conferencing. It uses WebSockets for low-latency messaging and WebRTC for peer-to-peer media streaming. This pushed me to think about persistent connections, real-time state synchronization, and building responsive systems under continuous load.
After that, I built create-reactgo-app, focusing on developer experience. I was frustrated with repetitive setup, so I created a CLI that scaffolds a containerized React and Go app with a single command. It ended up being used by 400+ developers, which was my first real signal that something I built was genuinely useful.
Most recently, I built Lumen, where I focused on performance engineering and distributed systems. I designed a Go-based probing engine for network-level tracing and high-concurrency load testing, capturing detailed metrics like DNS, TCP, TLS, and TTFB. This project pushed me to think more deeply about concurrency, system limits, and how real-world APIs behave under load.
four years of late nights, group projects, and learning how to ask better questions
I am a generalist at heart who can build with anything, but here is the core stack I have spent the most time with. Hover the logos, they blush a little.
I publish my thoughts on Medium instead of building a custom blog. Rather than reinventing infrastructure, I choose a mature platform that lets me focus on what truly matters: sharing practical insights on new technology, product strategy, and scalable technical architecture. No CMS drama, just write and ship.
dog-eared pages and tabs I keep coming back to.
Beyond building software, I approach engineering as a system of decisions that balance technical rigor with real human needs. I am driven by curiosity whether that means diving deep into distributed systems, refining product architecture, or understanding how a solution works at its core.
me, somewhere between a deploy and a coffee refill
Outside of engineering, I value reflection and real-world perspective. I believe the best products are built by people who explore broadly and think deeply. That balance between technical correctness and human awareness shapes how I design systems: scalable, intentional, and built to resonate with real users.
Connect with me on LinkedIn or . I reply faster than my CI pipelines.
no forms, no calendars required. a short note is enough.
thanks for stopping by
You made it to the end. Before you go, steal a quiet focus snack, or just sit with the clock for a minute.
focus · a little timer for whatever's next