Chief Technology Officer

CTO - UpToCode (uptocode.ai)

AI-Powered Construction Compliance \| Equity / Vesting \| Remote (Australia preferred)

About UpToCode

We’ve built an AI platform that reads construction drawings and BIM models, then checks them against building codes and regulations in minutes - work that currently takes building surveyors \& Architects/Engineers 4-6 hours per project. The product is live, the architecture is production-grade, and we’re onboarding our first paying users now for the pilot.

We’re looking for a highly technical CTO to own the platform end-to-end and scale it from working product to market leader.

What’s already built:

The backend is a FastAPI service running a multi-stage AI pipeline. PDFs are processed through hybrid chunking for document extraction, BIM files are parsed via an openshell into structured building data (walls, doors, windows, sill heights, materials - with relationship traversal and multi-step property fallbacks). Everything is embedded using Gemini and stored in PostgreSQL with pgvector for vector similarity search.

The query engine is a 4-node LangGraph state machine (rewrite → retrieve → classify → answer) with a multi-round tool-calling loop - 4 tools, 20\+ query actions, 15 filter parameters, and auto-escalation between Claude Sonnet and Opus based on query complexity. Full compliance reports cross-reference drawings, IFC data, regulation documents, and live web search across 50\+ tool rounds. Background processing runs on Redis \+ RQ with two async queues, crash recovery, and a full audit trail.

The frontend is Next.js. Infrastructure is Docker Compose deploying to Railway with PostgreSQL \+ pgvector.

What you’d own:

The entire technical roadmap. Scaling the RAG pipeline for accuracy and speed. Optimising LLM costs across Anthropic and Google models and potentially replacing with a sovereign ai model we are in the process of aquiring and tuning. Building out the team. Expanding BIM parsing beyond ArchiCAD to Revit, Tekla, and other authoring tools. Improving retrieval quality - embeddings, chunking strategies, cosine thresholds. Adding jurisdiction-specific compliance modules. Architecting for multi-tenancy, SOC 2, and enterprise readiness.

What we’re looking for:

Deep hands-on experience with LLM application development - RAG pipelines, embeddings, vector databases, prompt engineering, multi-agent orchestration, tool-calling architectures. Not someone who’s read about it. Someone who’s shipped it.

Strong Python backend skills - FastAPI or equivalent, PostgreSQL, Redis, async processing patterns. You should be comfortable reading the architecture doc and immediately seeing what you’d improve.

Experience with at least one of: construction technology, BIM/IFC, document processing (OCR, PDF extraction), or regulated industries where compliance matters.

A track record that investors can point to. Prior startup experience, a successful exit, a product that scaled, or meaningful open source contributions. We’re raising soon and need a CTO whose background gives investors confidence.

The deal:

This is a co-founder/CTO role with vesting equity, not a salaried position. We’re looking for someone who wants to build something significant in an industry that’s barely been touched by technology - construction productivity has grown 0\.4% annually for 80 years. The compliance software market is $1\.5-3 billion and growing at double digits. We have a working product, early users, and a clear path to revenue. We need a technical partner to turn it into a company who will join us on the journey to going public.

Our stack at a glance:

Python, FastAPI, PostgreSQL \+ pgvector, Redis \+ RQ, LangGraph, Anthropic Claude (Opus \+ Sonnet), Google Gemini (embeddings \+ vision), ifcopenshell, Tavily, Docker, Railway, Next.js, SSE streaming

If this is you: send a message with what you’d change about the architecture described above. That tells us more than a CV but also maybe include your acheivements...