Hello, I'm Mayank Raj. I build infrastructure, security, and data systems that are expected to keep their composure when production starts throwing furniture. I also play the violin, trek whenever the mountains allow it, and occasionally convince drones to behave. Mostly.
By day, I'm a Staff Engineer at Stripe on the Core Infra team. That means I spend a lot of time thinking about the boring-looking foundations that quietly decide whether everything above them feels fast, safe, and reliable. The glamorous part is architecture. The honest part is making sure the floor does not wobble.
Before Stripe, my work moved across security, cryptography, big data, AI systems, and cloud platforms. Different rooms, same obsession: make the complex thing understandable enough that teams can operate it without needing a campfire story and three tribal elders.
- At Salesforce, I worked on security and cryptography for large-scale enterprise systems, including benchmarking, caching, and reliability work that improved throughput while cutting cost. Very glamorous if your idea of glamour includes threat models and suspicious latency graphs.
- At Cactus Labs, I led a 15+ engineer team building big data and machine learning systems. We built a data platform that handled 1.5TB of data per week and turned it into decisions people could actually use, which is the part the architecture diagrams often forget.
- As the founder of Apptale.io, I created a monitoring service that watched systems across all 51 AWS availability zones at roughly 1/15th the cost of comparable tools. Infrastructure tends to whisper before it screams. Apptale was built to hear the whisper.
- I've also contributed to open-source projects like FireFox. Yes, that browser you might be using right now. Tiny fingerprints, but fingerprints nonetheless.
When I'm not deep in the machinery, you might find me scaling mountains, coaxing melodies from my violin, or tinkering with drones. The through-line is the same: systems are easier to understand when you respect both the theory and the messy physical world they live in.
The work I enjoy most sits at the uncomfortable boundary between architecture and operations. The diagram says one thing. Production, with its charming lack of respect for diagrams, says another.
That is where I like to work: reliability problems, security boundaries, data systems, cost cliffs, and the small design choices that quietly decide whether a system is pleasant to operate or a permanent group project with incident management.
A few representative scars:
- I've architected a graph database engine that can query ~8TB of raw data for under $30 per query, with an average query time of 8 minutes. That is less "needle in a haystack" and more "the haystack is on fire, the needle is invisible, and finance still wants the query to be cheap."
- I've spoken at conferences, hosted OpenAI Codex community meetups, judged and organized hackathons, and built an AR bot assisted by LLMs for Salesforce conferences, used by over 450 unique users across 4 conferences. Systems are fun. Rooms full of curious builders are better.
- I also run Sudomeet, a place for builders to meet, share what they are learning, and make the lonely parts of technical growth a little less lonely. Community work is still infrastructure. The packets are just people.
If you want to talk about infrastructure, reliability, security, cloud architecture, AI systems, trekking routes, or why the violin is basically distributed systems with nicer failure modes, drop me a line!


