Navigating the
Idea Maze
We spent the last year navigating the idea maze. We've worked through eight different product iterations and learned so much. Today marks the one-year anniversary of the incorporation of Tools for Sovereignty. This is an account of what we worked on, what we found interesting, and what we learned. PostHog and Lenny's newsletter, which also influenced our navigation of this phase, are valuable resources for any early-stage founder pre-PMF.
The Startup Company
Product: Incorporation + Neobank using ZKPs (Aztec) — think Stripe Atlas + Mercury.
Problem: Admin is one of the blockers preventing entrepreneurship outside the US.
Solution: Simplest way to incorporate & run an internet native company.
GTM: Partner with Sark, a UK crown dependency, to create a hub for internet founders.
What Did You Get Done Today?
Product: A Slack bot automating standup.
Problem: Companies waste a lot of time on coordination whilst all the work artifacts already exist.
Description: A Slack bot using LLMs to aggregate the work done across different tools.
GTM: Founder sales.
Personal Data Marketplace
Product: Marketplace for internet users to sell their data to train models.
Problem: Get model providers to pay internet users for their data & get people to care about data ownership.
GTM: Bootstrap demand with crypto economics.
The Worldcoin Bet
Product: Mini-apps on World App.
Opportunity: World App was emerging as a novel superapp.
GTM: Build the best AI apps on World using personalisation — got to 250k users in the span of a few months.
Private Memory Infrastructure
Product: Private memory infrastructure sold to AI apps.
Problem: With our apps we realised early that memory was the moat for AI apps — users should control it and be able to share it across apps.
Solution: SDK/API & MCP server for apps to personalise their LLM-based offerings, privacy guarantees via TEEs.
We wrote a manifesto about why this mattered.
The Identity Play
Product: IAM provider with context seeding.
Problem: Users don't have to reintroduce themselves to every app.
Solution: Identity provider with context seeding — think Poke's onboarding as a service.
Fast Iterations
Products: MCP context aggregator, initially for individuals, then for companies.
Problem: Your context is tied to an LLM provider, you should be able to port it.
Solution: An extension to your AI apps that syncs them to the relevant context.
Coworker in Slack
Product: Agent that lives in Slack and gets work done for you.
Problem: Most agents today are single player, require technical knowledge to adopt and aren't proactive.
Solution: A digital coworker that knows your company, adapts to how you work and is accessible directly in your team communication tool.
We're a small team of founders obsessed with product and sweating the details.