AI companion spin-off

Context

Cleevio, a Prague agency, spun up a B2C AI companion app in early 2024 - among the first teams shipping ChatGPT plus voice (ElevenLabs) in one product.

Problem

Prove a multimodal companion PoC fast enough to raise pre-seed, with no established playbook for UX, safety, or monetisation.

My role

Product Manager: research, PRDs, investor narrative, and coordination with design and engineering through PoC and funding.

Constraints

• Speed: Hype window and investor interest were time-boxed.
• Ethics and JTBD: Companion positioning vs tool use was unresolved and uncomfortable for me over time.
• 0→1: Product, pitch, and spin-out in one thread.

What I did

• Ran user interviews and market research; defined requirements for chat + voice experience.
• Shipped a working PoC with engineering; built materials for the investor story.
• Supported the spin-out and CZK 7.5M pre-seed round with research and PRDs.

Key decisions

• Ship PoC before perfect policy: Learn in market with a fundable demo.
• Step off after the round: Work succeeded on its terms; I did not want to continue on a companion trajectory I did not believe in.

Result

PoC shipped; company spun out with pre-seed funding. Early traction through launch (not thousands of DAU claimed here - the win was fundability and ship speed). More context on the Cleevio chapter.

What this shows

0→1 product under ambiguity: research, requirements, and funding narrative for an AI-enabled product - then a deliberate exit when the job-to-be-done did not match how I want to build.