Make AI memory correctable, explainable, and erasable.
AI products are getting long memories. Lineleaf exists so that every one of those memories can be traced to its source, corrected when the source changes, and removed — with proof — when it has to go.
Where we are
Lineleaf is at the very beginning — there is no released product yet. Here is what is true today.
- Early developmentin progress
- Domain secureddone
- GitHub organization createddone
- Python and npm package names reserveddone
- Initial product concept defineddone
- Looking for design partnersin progress
- MVP development beginningin progress
Roadmap
Ordered, not dated. We will add dates when we can keep them.
- Design-partner conversations
- Landing page
- Architecture validation
- SDK
- Lineage graph
- First adapters
- Dry-run invalidation
- Receipt prototype
- Open-source release
- Managed cloud
- Audit exports
- Team access
- Enterprise deployment options
Building AI products with memory?
Design partners shape the first adapters, the invalidation workflow, and the receipt format. If derived copies of user data worry you, we would like to hear how your stack works today.
Become a design partner- hello@lineleaf.dev
- GitHub
- github.com/lineleaf-ai
- Web
- lineleaf.dev