About Convore
Convore is a developer-focused communication and email integration product for testing and quality checks on conversational AI.
Who it's for
Developers and small teams who build chat and voice agents. That includes indie developers shipping their first assistant, startup teams adding one to their product, and SaaS teams who need to show what was checked before a release. Software engineers and AI/ML engineers use it alike, mainly in English- and Spanish-speaking markets.
Why it exists
Too many agents get tested by their users. The bad answer, the wrong tone and the policy slip show up in production, in front of the people you built it for. Small teams rarely have QA staff, and most testing platforms sit behind demos, trials and enterprise contracts.
Testing a conversation should be as normal as testing code. Convore exists to make that the default:
Test your AI conversations before your users do.
What it does today
Convore follows one loop: plan the scenarios, review the conversations turn by turn, then check every prompt change for what needs re-testing. Each step has a free tool:
- Scenario Builder turns one line about an agent into up to 10 test scenarios with pass and fail lines and a checklist, exported as Markdown or JSON.
- Transcript Review Scorecard splits a transcript into turns, collects your 1–5 scores for correctness, tone and policy safety, flags weak turns and exports a summary.
- Prompt Diff Check compares two system prompts line by line and word by word, then builds a re-test checklist.
Email integration is part of Convore. More details are coming.
How this site works
Where it's heading
Every team that ships a chat or voice agent tests the conversation before users see it. Scenario lists, scorecards and prompt diffs sit next to the prompt in the repo and are re-run on every change. See what comes next on the Roadmap & Pricing page.
Questions, ideas or problems? Contact us, or write to info@convore.dev.
