Foundgine
End-to-end evidence · Run 1 · 17 August 2026

Establish the baseline: can semantic execution reduce agent work?

Run 1 starts with the smallest useful experiment: the same customer-exposure-review business process is completed through a conventional application/tool flow and through Foundgine. The purpose is not to prove speed. It is to establish whether the semantic path can compress the agent interaction while reaching exactly the same final state.

Inspect the matrix evidence →All five runs
Why this run existsThis is the control experiment. If the semantic boundary cannot reduce the number of application interactions or context carried between them, the later scalability and efficiency tests have little to build on.
What was comparedThe same customer-exposure-review business process through conventional and semantic agent flows.
What it meansFoundgine was already changing the shape of the agent interaction before optimization: fewer application-level steps and less estimated context, but with execution overhead that later runs needed to investigate.
Published finding
What the evidence says

The semantic flow reduced tool calls from 7 to 4 and estimated context load from 932 to 534 tokens (42.7% lower), while reaching the same final state. The trade-off was higher wall-clock time: 12.69 ms conventional versus 24.25 ms Foundgine (+91.1%). The early result therefore supports an agent-work reduction, not a latency claim.

Key measurements
toolCalls7 → 4
context932 → 534 est. tokens
wall12.69 → 24.25 ms
sameStateTrue
How to use this run

Read the finding first, then inspect the exact evidence.

This page tells the story of the experiment. The benchmark explorer contains the interactive workload/concurrency matrix and selected-cell comparison.

The benchmark distinguishes measured timings and tool calls from its estimated context-load heuristic.

← Benchmark landing page · How it works · Architecture