Philo Ventures is a venture studio based in Utah that supports founder-led companies. They are seeking a Technical Co-Founder for Covellum, an AI-powered pre-construction cost intelligence platform, to own the product and technology end-to-end, transitioning into a CTO role as the company scales.
Responsibilities:
- Own the technical and product vision. Make the core calls on stack, data model, agent orchestration, retrieval strategy, and infrastructure — and on what the product actually is. Decide what to build, what to buy, and what to defer. You are the architect of record and the product mind behind the system
- Design the agentic core. Covellum is an agentic product. You will design how agents ingest plans, query cost data, reason about scope and assumptions, and produce defensible estimates — including the orchestration, tool use, memory, and evaluation layers underneath
- Build the data layer. Construction cost intelligence is a data problem before it's an AI problem. You will design the pipelines, schemas, and storage (relational, vector, geospatial) that let Covellum produce localized, accurate estimates and improve over time
- Ship the 0-to-1 product. Drive the build to first paying customers. Philo's engineering team builds alongside you — you are not the lone engineer trying to ship everything yourself. As Covellum scales, transition from primary builder to engineering leader without losing technical depth
- Translate customer reality into the system. Sit in design partner calls. Understand what an architect, a developer, or a construction lender actually does with cost data, and translate that into a product they trust. You are the technical voice in every customer conversation and the customer voice in every technical conversation
- Build the engineering team. Recruit the founding engineers when the time is right. Set the bar for technical hiring, code quality, and how the team works
- Drive technical strategy. Stay ahead of the model, infrastructure, and tooling landscape. Make sure Covellum is building durable advantages, not just riding the current wave of AI capability