How it works

From git history to a filing-ready claim

SREDify pairs AI agents with a structured claim workflow. The AI proposes, and you decide.

Step 1

Connect sources

Link repositories read-only and upload technical docs, timesheets and prior claims.

Step 2

AI agents mine your history and surface the work that matters, with links back to the source.

Step 3

Accept, merge or reject suggested projects. Answer targeted questions to fill the gaps.

Step 4

People, time and costs roll up into expenditures and estimated investment tax credits.

Step 5

Generate T661 narratives, edit them, and run an AI review before you file.

Step 6

Download a filing-ready package with descriptions, financials and supporting evidence.

The AI agents

Reads commits, pull requests and issues across GitHub, GitLab and Bitbucket, and triages what looks like experimental work.

Groups evidence into candidate SR&ED projects, flags routine engineering, and asks the questions that frame your uncertainties.

Drafts lines 242, 244 and 246 within CRA word limits, grounded in your evidence, answers and guidance.

Checks drafts and numbers the way a reviewer would: missing uncertainties, weak advancements, unsupported claims.

Connect a repository and walk through the workflow for free.