Thredmark

Thredmark.Memory that shows its sources.

Thredmark is source-backed working memory for evolving AI workflows, with a Python API and paired evaluation tools.

Thredmark · deal-review workflowSL
Working memory
142 itemsPython API
FactSourceVersionUpdated
Renewal is annual, 30 days noticecontract.pdf · p4v22h ago
Customer prefers weekly deliverycall notes · 12 Marv13d ago
QA bar raised to 95% agreementspec-v3.md · §2v31h ago
Pilot covers oncology onlyemail · 9 Marv15d ago
Data must stay in EU regiondpa.pdf · §7v16d ago
contract.pdf · page 4

…this Agreement renews annually unless either party gives 30 days written notice before the renewal date…

Evaluation run
RecallGrounded in sourceStale facts replaced
Illustrative interface

What it does

Long-running AI workflows need to remember what they learned along the way, and that knowledge changes. Thredmark keeps a working memory where every item is tied to its source, so an agent can show where a fact came from and update it when the source changes.

Python API and evaluation tools

Thredmark comes with a Python API to add to and query the memory from your workflow, and paired evaluation tools to test what it recalls, before and after changes.

Every safe model
has an expert behind it.