| What it does | Managed AI trainers, LLM evaluators and benchmark authors for AI labs and data partners
| The night watch for apps built with AI that tests checkout on every preview, watches live apps and payments, and identifies the deploy behind a break
|
|---|
| Stage | Not given
| Not given
|
|---|
| Founded | Not given
| Not given
|
|---|
| Based in | IN
| IN
|
|---|
| Pricing model | Contact for pricing
| Freemium
|
|---|
| Pricing | Not given
| Watch: Free (1 app, 1 journey read-only on production daily, security checks, 1 uptime monitor). Guard: $19/month (5 journeys on every preview and production deploy, business pulse, 20 auto-fixes a month, 10 monitors). Guard + Calls: $49/month (everything in Guard plus 10 journeys per app, Corvis ca
|
|---|
| Key features | - Coding-agent benchmarks in Terminal-Bench and SWE-bench style
- Research reproduction in PaperBench style
- Model evaluation and RLHF with preference ranking
- Long-horizon agent tasks with planning and tool use
- STEM and expert reasoning questions at PhD level
- Trial task scoring before work begins
- Peer review on every item
- Senior reviewer calibration on each batch
- Product design and full-stack development for SaaS and AI
- LLM assistants and voice AI integration
| - Journey testing on preview deployments in real browser
- Security checks on every push
- Uptime monitoring
- Leaked key detection and rotation help
- Screenshot and fix brief for failing steps
- Telegram alerts with voice briefings
- Export journeys as Playwright tests
- Supabase row-level security check and auto-fix
|
|---|
| What makes it different | - One programme lead and point of contact instead of managing hundreds of individual profiles
- Scored trial on your rubric for every expert, not just CV and quiz screens
- Second review on every item plus senior calibration, not just spot checks
- Published quality numbers: 2,500+ tasks authored, 90% acceptance rate, 4.89/5 rating
- Pilot staffing in days with 5 to 10 experts
| - Tests real customer journeys in a real browser on every preview, not AI-generated tests grading their own work
- Learns your app's normal payment and signup patterns to detect drops when they occur
- Works across Vercel, Supabase and Stripe together for full app visibility
- Calls you on critical issues like stopped payments instead of just sending alerts
|
|---|
| Community votes | 0 | 0 |
|---|
| DR (Domain Rating by Ahrefs) | 2 ▼ −3 (down) since last check | Not checked yet ● no change since last check |
|---|
| Trust Flow (Majestic) | 0 | 0 |
|---|
| Citation Flow (Majestic) | 0 | 0 |
|---|
| Referring domains (Majestic) | 1 | 0 |
|---|
FynoLabs
Corvis