Teams lose days synthesising analytics
Finding why users drop off means piecing together fragmented data across tools.
Make AI customers test before users do.
Fraser autonomously tests your product with AI customers, turns feedback into fixes, and continuously improves your product, all within your team’s existing workflow.
@Canaan created this channel. This is the very beginning of the # general channel. (Edit description)
Today
Trusted by
teams at:
teams at:
Product decisions move faster than customer feedback can keep up.
Customer friction goes undetected
Too many users hit broken or confusing experiences before teams can respond.
Create AI personas from
real customer data
User
PM
Labelling user dataAI Vision & Audio
Turn interviews into AI personas
Transform user interviews into AI personas grounded in real customer conversations.
PostHogread-only
Sessions read0
#4821exit · plan picker0:42
#4822rage-click · Continue1:07
#4823exit · signup0:19
#4824back · plan picker2:15
#4825idle 30s · pricing1:53
#4826exit · empty state0:08
#4827rage-click · Compare0:57
Enrich AI personas with context
Feed AI personas real user behavior from tools like PostHog.
Patricia’s run3 filedAdd to basket does nothing
LYT-482screenshot · Pixel 8
Keyboard covers Continue
CHK-119screenshot · iPhone SE
Skip sits below the fold
#214screenshot · Web
You approve every ticket. Fraser has no code access.
Turn findings into tickets
Send issues directly to Jira, Linear, Github with context attached.
Why teams use simulations
What do my customers think of my new revamp?
iPhone users · US95% confidence
Confidence level95%
LowHigh
Validate design choices
Find friction in user flows before launch.
!Missing
e.preventDefault() — form conflict on mobile. // CheckoutButton.tsx
−onClick={() => handleSubmit()}
+onClick={(e) => { e.preventDefault(); handleSubmit(); }}
disabled={isLoading}
Fix applied
Catch bugs
Detect broken features before production.
👩🏻Sarah
🧑🏽Tyler
👩🏾Maya
🧑🏼Jordan
👨🏿Alex
Improve UX
See where users get confused before you ship.
👩🏻
SIM-Maya-Writer-01
I don't like the wording on landing.
What do you not like about it?
👩🏻
SIM-Maya-Writer-01
Pretty much, the use of "was" is everywhere.
Persona Debrief
Follow up with AI personas to understand the rationale behind every action.
A
Authenticated states
Test flows behind login, paywalls, and role gates.
LoginAdmin
M
Multi-step journeys
Walk through onboarding, checkout, and complex forms.
OnboardingCheckout
Navigate complex product flows
Test the pages, states, and edge cases that ordinary agents cannot access.
MCP Server
From feedback to fix
Turn user issues into prioritized fixes your team can approve and build.
Built with teams who test.
Matt Straub
Customer Success @ Limitless
“What stood out was seeing the onboarding issues we could fix before customers ever encountered them.”
“I can put a new flow in front of Fraser and come away knowing exactly what still feels unclear.”
Roli Srivastava
Engineering Leader @ Amazon
“It fits naturally into pre-release testing, one more check that the experience is ready to ship.”
“Fraser could make user testing more continuous. It can simulate different types of users and help spot friction much earlier in the process.”
Pranshu Shekhar
Product Manager @ Walmart
“Fraser is good for teams that ship fast, saving time by streamlining pre-deployment testing and catching issues before release.”