$ curl -s sergedevs.com/whoami
HTTP/2 200 · content-type: application/json
{
"name": "Serge",
"role": "Inference & distributed systems engineer",
"experience": "10+ years",
"projects": [
7 items
]
}
Real-time AI,
measured in
milliseconds.
I build the infrastructure behind real-time AI — high-throughput event pipelines and prediction services measured in milliseconds. Strong Rust and distributed-systems background.
$ ps --projects
7 processes · 1 running full-time
01
Albatross
Lead Software Engineer
Feb 2025 — present
● RUNNING
Event ingestion and the low-latency prediction service behind real-time product discovery for some of Europe’s largest marketplaces.
02
PrepForIELTS
Founder
Web
● LIVE
open ↗
IELTS Writing Task 2 feedback in about a minute — scored against the four official criteria, with sentence-level fixes and re-scoring.
03
Clipset
Founder
macOS · Windows
● LIVE
open ↗
Desktop app that renders every combination of hooks, bodies and CTAs into separate video ad variations — locally, in minutes.
04
EchoRead
Founder
iOS
● LIVE
open ↗
AI language coach that turns a goal into a roadmap of stories, flashcards and an in-context tutor across ~35 languages.
05
wren
Founder
macOS · iPhone
● LIVE
open ↗
Menubar task list that shows five things. One keystroke to capture, everything else waits quietly until it is actually time.
06
Prompter
Founder
iOS
● LIVE
open ↗
Teleprompter for iPhone. Paste a script, read it as cards or let it scroll at your pace while you film on your own camera.
07
InferWorks
Founder
Consultancy
● ACCEPTING
open ↗
Inference cost-engineering for AI startups. Cut inference bills 10–50× — without trading away latency or quality.
$ ./contact --open