idappt is a foundation in pre-incubation stage and an AI-native engineering training lab that orchestrates autonomous AI coding agent networks. It architects and simulates multi-agent software "startups," structuring specialized artificial teams that include analysts, architects, engineers and testers.
idappt is a learning environment for its founders. We continuously develop our ability to work with increasingly capable AI systems — from context engineering and agent orchestration to autonomous software development and multi-agent collaboration.
We apply these capabilities to real-world problems. Human founders direct and collaborate with AI agents to research problems, design solutions and build practical applications that can reduce friction, save time and create societal value.
We use the lab to experiment with AI-native organizational models in which human founders and autonomous agents work as integrated teams. Through concepts such as compound startups, 20x companies, autonomous agent teams and artificial founders, we explore how companies, institutions and software development itself may evolve.
idappt engineers the context, orchestrates the models, and runs the agents. We design and operate end-to-end AI software "startups" — specialised teams of analysts, architects, engineers, and testers — led by a small group of humans whose judgment and accountability are behind every release.
Citizen headache, in numbers
452,879,405 email-address records exposed in verified data breaches during the last 12 months
Data citizens never chose to hand over, leaked by the companies and government agencies they have to trust.
Source: Have I Been Pwned · rolling 12 months by breach date · excludes spam lists, malware and stealer-log datasets · counts breach records, not unique people
Citizens have to hand over personal data to every service they use, and have no control over what happens to it afterwards. Breaches are the symptom.
Prototyping a new concept for a blockchain-based identity management and application platform, based on the Self-sovereign Identity paradigm.
Solidity contracts, React, JavaScript, TypeScript, Ethereum Sepolia, Moonbase Alpha (retired), Axelar GMP.
idappt decitizen is an experimental demonstrator under technical testing. Five Groth16 zero-knowledge circuits and their backend verification path are implemented and merged using synthetic test data and test-only setup material. Contracts were previously deployed and source-verified on Sepolia and Moonbase Alpha. The cross-chain pipeline is not currently operational following Moonbase Alpha’s recent retirement, and further security and reliability work remains.
TouchFire originated from an artificial founder exploring a simple question: what if achieving a goal depended less on motivation and more on the structure surrounding the commitment?
The concept combines commitment contracts, financial stakes, automated verification and small-group accountability to make following through easier than quietly giving up.
People often know what they want to do but still struggle to follow through. Motivation fades, habit trackers largely depend on self-reporting, and personal commitments often carry no immediate consequence when they are abandoned.
A mobile-first accountability app that combines real financial stakes, AI-assisted verification and small-group accountability — exploring whether stronger commitment mechanisms can help people bridge the gap between intention and action.
Foundation done, core business logic in progress.
AI-powered, mobile-first social-discovery app: users show where they are, share photos, video stories and short posts, and discover hyperlocal "hidden gems". They can also "talk" with SpotLike about the place they are in.
MVP under review.
