Tethys · founder and product engineer
Shape the problem, design the system, build the product, and stay responsible after release.
Founder · product engineer
Cedrik I build software that survives contact with reality.
Fifteen years turning messy operations into products, automations, and infrastructure people can actually run.
Ignition sequence
Scroll is the throttle.
The camera leaves the launch field here. The portfolio remains readable even when the flight layer is unavailable.
Builder is the role. Code is one of the tools.
I work where product judgment, engineering depth, and operating responsibility overlap. The useful question is not whether AI wrote a line. It is whether the system solves the right problem, ships, and keeps working.
I have spent fifteen years building across product, backend, mobile, automation, and infrastructure. That range matters because real systems fail at the seams.
I use AI aggressively. I also own the architecture, the tradeoffs, the production failures, and the conversation with the person whose day the software changes.
How I turn that into a delivery method →From messy process to operating system.
The through-line is ownership: discover the bottleneck, ship the narrow proof, and keep responsibility when real usage begins.
Rapidex · operating-system delivery
Turn WhatsApp conversations and voice notes into work that can be assigned, tracked, and recovered.
ClinicOS · clinic operations platform
Deliver human-in-the-loop automation for sensitive clinic intake, scheduling, and follow-up.
Rhapsody Integration Engine · healthcare interoperability
Product engineering for the integration engine that moves more than a billion clinical messages a day across 1,900+ healthcare organizations in 31 countries.
UniCredit · CEE Digital Channels
Frontend, backend, and production support for the internet banking of Central and Eastern Europe's largest banking group — 25 million clients across 13 markets.
SS&C Intralinks · virtual data rooms
Full-stack and reliability engineering for the virtual-data-room pioneer trusted by over 95% of the Fortune 500, with $35+ trillion executed on platform.
Product judgment, engineering, operations.
Product clarity, scope, adoption, and operations decide whether a build becomes leverage or another abandoned repository.
- 01
Find the operational truth
Start with the bottleneck, the person doing the work, and the decision that keeps getting delayed.
- 02
Ship the smallest proof
A narrow system in real use beats a broad platform waiting for permission to matter.
- 03
Own the boring end
Deployment, observability, handoffs, and recovery are part of the product, not cleanup after it.
Work with fingerprints on it.
Each project is a real operating system, product, or public experiment. The code supports the story; it is not the story.
01 Live
Product studio
Tethys
Productized software, AI workflows, and internal systems built around measurable operational outcomes.
02 Live
Logistics SaaS
Melho
A logistics platform for local operators: order intake, pricing, PIX payments, dispatch, route optimization, and live tracking.
03 In production
Logistics operations
Rapidex
A WhatsApp-first operating system that turns voice notes and conversations into trackable delivery work.
04 Delivered
Clinic operations
ClinicOS
AI-assisted intake, scheduling, and follow-up designed to hand sensitive decisions back to clinic staff.
05 Delivered
Restaurant operations
Aska Takeout
A complete takeout operation for a ramen house: WhatsApp ordering agent, live pickup queue, payments, and customer feedback.
06 Validating
Automotive consumer
Deviz Corect
Human-reviewed audits of Romanian car-repair estimates: parts pricing, labor-overlap checks, and negotiation guidance.
07 Publishing
Build in public
Piemental
The honest record of building products, making decisions, and showing the work without guru theatre.
04 / Mission control
Got an ugly operation? Good.
Send me the bottleneck, who feels it, and what a useful outcome looks like. I will tell you plainly whether software should touch it.