SpeqForgeGo from customer feedback to ready-to-build specs, no product manager required.
SpeqForge generates complete user stories, acceptance criteria, and effort estimates, all grounded in real customer evidence, ready in seconds.
Example auto-generated spec
As a data analyst, I want to export my project results as a CSV file so that I can build custom reports in my existing spreadsheet tools without manual copy-pasting.
- Given I am on the results page, when I click "Export CSV", then a download starts within 2 seconds
- Given the results set is empty, when I click export, then I see a message "No data to export"
- Given the file is generated, when it downloads, then it includes all columns visible on screen
- Given I have no export permission, when I attempt export, then I see a clear upgrade prompt
Show a spinner inside the Export button; disable the button and add "Generating…" text while the file is being built
If the result set is empty, hide the Export button entirely and show an inline message: "No data to export yet"
If generation fails, show a red inline error below the button with a "Try again" link. Do not navigate away from the page
Hide the Export button and show a locked icon with a link to the upgrade page
- ▸User triggers a second export before the first download completes: keep the button disabled until the download begins, then re-enable
- ▸Dataset exceeds 100,000 rows: warn user with a count and estimated time before starting
- ▸Column header contains a reserved Excel character (#, @): allow export but append a note to the confirmation message
- ▸User is on a read-only share link: do not show the Export button at all, not even disabled
Generate the CSV server-side to handle datasets above 10,000 rows. Stream the file to the browser using Content-Disposition: attachment rather than storing it on disk. Escape all values per RFC 4180, including fields containing commas, quotes, or newlines.
Specs link to behaviours and scenarios
How it works
Select a feedback theme
Pick any clustered theme from your Feedback Analysis, or describe a feature you want to build from scratch.
Spec auto-generated
Drafts a full user story, acceptance criteria, edge cases, and effort estimate grounded in your actual customer evidence.
Share or export
Lands as a work item on your SpeqForge workboard, or share a read-only link with stakeholders. No formatting required.
Everything in a spec
All the context your engineering team needs, in one card.
"As a [user], I want [goal] so that [benefit]", correctly structured every time.
Given/When/Then style criteria that engineers can immediately use to define done.
S/M/L/XL estimates inferred from feature complexity and common engineering patterns.
Customer quotes are embedded in the spec so you know why the feature is being built.
Flags error states, empty states, and permission scenarios you might have missed.
Every spec gets a public read-only link, so you can send it to stakeholders without account access.
Before vs. after
Where this fits in your workflow
Frequently asked questions
Can I edit the spec after it's generated?
Yes. Every spec is fully editable in the app before you export. You can adjust the user story, add or remove criteria, change the effort estimate, or add technical context. The generated spec is a structured starting point, not a locked final draft.
Does it work if I don't have feedback data?
Yes. You can describe a feature in plain text and SpeqForge will generate a spec from that description alone. Feedback Analysis integration adds customer evidence and better prioritisation, but it is not required, and the spec generator works standalone.
How does it handle complex multi-part features?
For larger features, SpeqForge will produce a parent spec and suggest breaking it into sub-stories with their own acceptance criteria. You can keep it as a single spec or promote the sub-stories into independent tickets with a click.
What's the difference between this and writing specs in ChatGPT?
SpeqForge generates specs grounded in your actual customer evidence, so the quotes and themes come from your real feedback data. It also structures output in a consistent format every time, handles edge cases from real user behaviour, and links directly to your project tracker.
A generic chatbot produces plausible-sounding specs with no customer evidence behind them.
Stop writing specs from scratch.
Generate your first evidence-backed feature spec from your own feedback. Free to start, no credit card required.
Get started free