The labThe way an
The way an
AI-native team works
A look inside SDEN: the technologies we run, the AI agents live in production, and how we think and operate: from the software to the people. The same path we take clients down.
The fleet, running
The agents we run in production.
Not a roadmap. The agents serving this site and the platform behind it, right now: runs per minute, models in use, success rate, latency, and the tools they reach for.
Live · 9 agents active · synced 7 d ago
/lab · sden.aiAgents in production
0
9 running right now
Runs · last 24h
0
~3.3/min sustained
Success rate
0.0%
Across all production agents
Latency p50
0 ms
p95 4200 ms
Orchestration
Agent fleet
Intake triage1,840 runsrunning
Site audit streamer1,120 runsrunning
Lead router760 runsrunning
Doc retriever (RAG)540 runsrunning
Prompt optimizer310 runsidle
Readiness scorer250 runsidle
Inference
Models in use
claude-haiku-4-571%
claude-sonnet-4-629%
Tokens · 24h18.4M in · 2.1M out
Capabilities
Tool calls · 24h
web_search9,120
retrieval6,340
code_exec2,110
send_email480
Est. compute cost$41.7
Figures are drawn from our own orchestration layer and production monitoring; the page is statically exported and animates client-side, so no request hits a server. Last sync: 7 d ago.