Follow the plan through your data. Every answer comes with its path and the decisions that kept it.
Request a demo
Swipe up to continue
Every node explains itself. Select any node to see which decisions kept it at each hop.
Request a demo
Swipe up to continue
Every candidate says why. Each node reached at a hop is scored, then kept or left out, with the reason.
Request a demo
Swipe up to continue
Hop by hop, through real data. The plan is walked through your graph one relation at a time, from the start node to the answers.
Request a demo
Swipe up to continue
Answers with their paths. Each answer node comes back with the exact path that reached it.
Request a demo
Swipe up to continue
Wayfinder plans. Waytracer computes the answer. The top-ranked route arrives ready to walk, with the alternatives kept alongside.
Request a demo
Swipe up to continue
More about Waytracer. What it takes to run it, and how it measures up.
Request a demo
Swipe up to continue
Built for production
Built for production
One hop at a time
Each hop is its own step you can inspect.
Results you can use
Answers, paths and reasons as JSON.
Benchmarks
Every top answer right, on our 82-question benchmark. 3–11× faster than frontier models, at $0.59–$3.85 per 1,000 queries.
| Benchmark | AiQL Wayfinder + Waytracer | Anthropic Opus 5.5 | Anthropic Sonnet 5.5 | OpenAI GPT-6.1 |
|---|---|---|---|---|
| Hetionet Top answer right, 42 questions | 100% | 100% | 92.9% | 95.2% |
| FHIR Top answer right, 40 questions | 100% | 72.5% | 70% | 75% |
| Hetionet Answer F1 | 1.00 | 0.93 | 0.87 | 0.87 |
| FHIR Answer F1 | 0.91 | 0.72 | 0.68 | 0.74 |
| Hetionet Median latency | 2.3 s | 12.1 s | 7.0 s | 11.5 s |
| FHIR Median latency | 2.8 s | 31.3 s | 24.0 s | 27.9 s |
| Hetionet Cost per 1,000 queries | $0.59 | $40.73 | $22.42 | $12.18 |
| FHIR Cost per 1,000 queries | $3.85 | $377.37 | $190.41 | $16.25 |
| Hetionet Failed runs | 0 | 0 | 3 | 2 |
| FHIR Failed runs | 0 | 8 | 12 | 10 |
Hetionet
Top answer right, 42 questions
AiQL Wayfinder + Waytracer100%
Anthropic Opus 5.5100%
Anthropic Sonnet 5.592.9%
OpenAI GPT-6.195.2%
FHIR
Top answer right, 40 questions
AiQL Wayfinder + Waytracer100%
Anthropic Opus 5.572.5%
Anthropic Sonnet 5.570%
OpenAI GPT-6.175%
Hetionet
Answer F1
AiQL Wayfinder + Waytracer1.00
Anthropic Opus 5.50.93
Anthropic Sonnet 5.50.87
OpenAI GPT-6.10.87
FHIR
Answer F1
AiQL Wayfinder + Waytracer0.91
Anthropic Opus 5.50.72
Anthropic Sonnet 5.50.68
OpenAI GPT-6.10.74
Hetionet
Median latency
AiQL Wayfinder + Waytracer2.3 s
Anthropic Opus 5.512.1 s
Anthropic Sonnet 5.57.0 s
OpenAI GPT-6.111.5 s
FHIR
Median latency
AiQL Wayfinder + Waytracer2.8 s
Anthropic Opus 5.531.3 s
Anthropic Sonnet 5.524.0 s
OpenAI GPT-6.127.9 s
Hetionet
Cost per 1,000 queries
AiQL Wayfinder + Waytracer$0.59
Anthropic Opus 5.5$40.73
Anthropic Sonnet 5.5$22.42
OpenAI GPT-6.1$12.18
FHIR
Cost per 1,000 queries
AiQL Wayfinder + Waytracer$3.85
Anthropic Opus 5.5$377.37
Anthropic Sonnet 5.5$190.41
OpenAI GPT-6.1$16.25
Hetionet
Failed runs
AiQL Wayfinder + Waytracer0
Anthropic Opus 5.50
Anthropic Sonnet 5.53
OpenAI GPT-6.12
FHIR
Failed runs
AiQL Wayfinder + Waytracer0
Anthropic Opus 5.58
Anthropic Sonnet 5.512
OpenAI GPT-6.110