FEATURE · ROI METER

Prove what your
agents are worth.

Send one event per task. Get back the hours and money saved, with every figure traceable to the baseline, rate and source behind it.

ALL AGENTS LIVE
  • contract-reviewer+€218.50
  • lead-enricher+€132.75
  • support-classifier+€64.00
  • invoice-extractor+€97.25
MONEY SAVED · THIS MONTH
€57,810
1284h SAVED · 7.3 FTE FREED
HOW IT WORKS

Log, resolve, price, prove.

01 · LOG

One call per task

Send agent, task type and outcome. Everything else is optional and defaults sensibly.

02 · RESOLVE

Baseline and rate

We resolve the human baseline for the task and your hourly rate, per agent, per org or fallback.

03 · PRICE

Auto-priced run

Pass model and tokens and the agent's own cost is priced against live model prices.

04 · PROVE

Net ROI, traceable

Hours, money and net saved come back, each figure carrying the source it resolved from.

EVERY FIGURE TRACEABLE

One call in. An auditable receipt out.

Pass model and tokens and the run is auto-priced, so the saving is net of what the agent itself cost. Finance can audit any single event.

track.shcurl
curl -X POST https://humanhours.dev/api/v1/track \
  -H "Authorization: Bearer hh_live_..." \
  -H "Content-Type: application/json" \
  -d '{
    "agent_id": "support-classifier",
    "task_type": "email_classification",
    "outcome": "success",
    "model": "claude-haiku-4-5",
    "tokens_in": 1800,
    "tokens_out": 240
  }'
RESOLVED EVENTevt_8aa6e7d8…
HOURS SAVEDbaseline 4m − run 0.06m0.066h
COST SAVED× €45/h€2.95
AGENT COSTclaude-haiku · 1.8k+0.24k tok€0.02
NET SAVEDcost − agent cost+€2.93
BASELINE 4 min · builtinRATE €45/h · org defaultPRICE auto · live model price
YOUR COCKPIT

Live tiles, a real trend, and a reports API.

TIME SAVED1,284h
COST SAVED€57,810
FTE EQUIVALENT7.30
SUCCESS RATE97.4%
NET SAVED · DAILY
Numbers above are illustrative. The same Bearer token pulls these into PowerBI, Looker, Excel or your own pipeline.

Ship one call. See your first proof.