You own engineering for the Unimatch Lab holding: technical leads, product CTOs, engineers, and infrastructure. You are the single technical gate for how the holding builds and ships.
This is a partner track, not a hired-hours role. You take the full technology domain, make it measurable, and run it without being told what to do next.
In the first year you set up the development system (planning, backlog, reporting, metrics), hire and shape the team and compensation model, set shared engineering standards across the portfolio, and take the AI-agent orchestration architecture to production.
You keep a helicopter view and stay hands-on: you know how to build and run multi-product development through cross-functional squads, and you can still ship a prototype, stand up Kubernetes, or debug a hard technical problem yourself.
Engineering system across the holding
- Own how the holding builds and ships: you are the single technical gate for technical leads, product CTOs, engineers, and infrastructure
- Stand up the development system for the first year: planning, backlog, reporting, and metrics, so delivery is transparent and forecastable
- Make the same workflow usable by any product in the portfolio, not a one-off process for a single team
- Keep dozens of parallel streams in control: prioritize, track, and close them to a result without losing the helicopter view
Team, org, and compensation
- Build the engineering organization from the current base: hire, structure squads and cross-functional teams, and set who owns what
- Design and run the compensation model for engineering so it matches holding scale, not a single-product startup
- Run daily management and tracking: what is in flight, what is blocked, what ships, what waits
- Shape the org so the team can grow without you managing every thread by hand
Standards and technical judgment
- Set shared engineering standards for all products in the portfolio and keep them in use, including how work is planned, reviewed, and shipped
- Decide architecture trade-offs at holding level: stack, patterns, constraints of hardware, memory, and model architecture, not only LLM tooling
- Unblock production problems yourself when the system needs it: prototype, Kubernetes, debug, ship
AI-agent orchestration
- Take the AI-agent orchestration architecture the holding runs on from design to production
- Own the engineering of agent pipelines, harnesses, and wrappers as a production system, not a research demo
- Experience building and running engineering at scale: tens of people and several products, including from scratch
- Deep AI background: LLMs, AI-pipeline orchestration, architectures and their limits (hardware, memory, transformer math, RL); background from DevOps through ML
- Full startup cycle: idea -> PMF -> rounds -> growth
- STEM or engineering degree
- Fluent English
- Can hold many parallel streams, work with large data, and still ship hands-on (prototype, Kubernetes, production debug)
- Organizational management, daily tracking, and hiring: strong engineers choose to work with you; you think in product and business, not only in code
- Big-tech experience
- Hardware and devices; deep work in TTS/STT and conversational AI
- Agent harnesses, wrappers, and orchestration of AI agents
- Startup path through IPO
- Top-tier university
- LLM, transformer architectures, RL
- AI-pipeline and AI-agent orchestration
- Constraints at hardware, memory, and architecture level
- Kubernetes and production infrastructure
- DevOps through ML
- A clear development workflow that any product in the portfolio can follow with predictable movement
- An engineering team that grows without day-to-day manual management from you
- The technical domain is no longer the bottleneck of the holding
You lead technical leads, product CTOs, engineers, and infrastructure across the holding. You are the single technical gate for build and ship.