Current technical landscape
The architecture is hybrid: AI video analysis runs server-side, the final clip assembly happens on-device. A strategic direction the CTO will own — gradually migrate decisions from backend to device without quality loss, to improve unit economics.
The voice stack is built on third-party providers (Whisper / Deepgram / OpenAI Realtime / ElevenLabs). The CTO must evaluate trade-offs between providers across latency, cost, and quality, and decide when to replace or hybridize parts of the stack.
The LLM agent (Operator / Director / Producer hierarchy) uses a memory system on pgvector + Neo4j (GraphRAG hybrid retrieval). The user portrait is built from multi-signal data — camera, voice, gallery, social analytics.
The engineering team today is 6 people, all reporting directly to the CTO:
•
AI / Agent: 2 people (including a recently hired AI Engineer)
Plus product, marketing, and growth. The plan for engineering team growth is targeted — +1–2 people over 6 months. This is not “build a team from scratch” — it is “lead an existing team and reinforce it where needed”.
Our product is entering a phase where the engineering team must become a full product engineering vertical with its own tempo, processes, and culture. We need a CTO who simultaneously owns three intersections:
•
Production mobile video stack at the level of Core Image / Metal / OpenGL ES / AVFoundation, video processing, rendering, on-device inference — real experience, not theory
•
Agentic AI and LLMs in production — LLM orchestration, RAG/GraphRAG, agentic architectures, memory systems, evaluation frameworks, hands-on with voice stacks
•
Subscription growth infrastructure at Codeway / Bending Spoons / Lightricks level — paywall A/B, MMP attribution, SKAN/AdAttributionKit, LTV/ROAS pipelines, cost optimization
This is the owner of the product engineering vertical, with direct accountability for the production product, hiring, unit economics, and technical strategy.