Acme Robotics — Production
⌘K
RK
Overview
What's happening across every AI workflow in this workspace, right now.
Last 14 days
All environments
New workflow
Runs, 24h
16,320
4.2% vs prior day
Overall success rate
94.2%
2.1 pts, driven by Lead Qualification
Avg latency (p50)
1033ms
— roughly flat
Spend, 24h
$1,392.20
Tracking 6% over monthly budget
Success rate across all workflows
Weighted by run volume · 14 days
100%
90%
80%
70%
SLO 80%
Aug 31
Sep 3
Sep 6
Sep 9
Sep 12
System health
8 workflows
8
workflows
4 Healthy
2 Warning
1 Critical
1 Paused
Workflow performance
Click a row to investigate
Ticket Triage & Routing
Support · Priya Nair
98.2%
Success
$128.40
Cost 24h
Invoice Reconciliation
Finance Ops · Marcus Webb
91.4%
Success
$412.10
Cost 24h
Contract Clause Extraction
Legal · Elena Torres
99.1%
Success
$96.20
Cost 24h
Sales Lead Qualification
RevOps · Priya Nair
76.3%
Success
$301.80
Cost 24h
Customer Churn Risk Summary
Customer Success · Devon Okafor
97.6%
Success
$54.30
Cost 24h
Refund Approval Agent
Finance Ops · Marcus Webb
94.8%
Success
$188.90
Cost 24h
Knowledge Base Answer Agent
Support · Priya Nair
96.9%
Success
$210.50
Cost 24h
Vendor Onboarding Doc Review
Procurement · Elena Torres
—
Success
—
Cost 24h
Needs attention
3 open
critical
Success rate fell below 80% threshold
Sales Lead Qualification · Today, 08:52
high
P95 latency exceeded 2.0s SLA
Invoice Reconciliation · Yesterday, 21:14
medium
Cost per run up 34% week-over-week
Refund Approval Agent · 2 days ago
Recent activity
Ticket Triage & Routing — Run tr_88213 completed
3 min ago
Invoice Reconciliation — Run tr_51032 exceeded SLA
8 min ago
Contract Clause Extraction — Run tr_22981 completed
14 min ago
Sales Lead Qualification — SLO breach confirmed
Today 08:52
Customer Churn Risk Summary — Run tr_16620 completed
22 min ago
Overview
What's happening across every AI workflow in this workspace, right now.
Last 14 days
All environments
New workflow
Runs, 24h
16,320
4.2% vs prior day
Overall success rate
94.2%
2.1 pts, driven by Lead Qualification
Avg latency (p50)
1033ms
— roughly flat
Spend, 24h
$1,392.20
Tracking 6% over monthly budget
Success rate across all workflows
Weighted by run volume · 14 days
100%
90%
80%
70%
SLO 80%
Aug 31
Sep 3
Sep 6
Sep 9
Sep 12
System health
8 workflows
8
workflows
4 Healthy
2 Warning
1 Critical
1 Paused
Workflow performance
Click a row to investigate
Needs attention
3 open
critical
Success rate fell below 80% threshold
Sales Lead Qualification · Today, 08:52
high
P95 latency exceeded 2.0s SLA
Invoice Reconciliation · Yesterday, 21:14
medium
Cost per run up 34% week-over-week
Refund Approval Agent · 2 days ago
Recent activity
Workflows
All Ai workflows in the this workspace, with live performance and cost.
New workflow
Workflow
Owner
Status
Runs, 24h
Success
Latency p50
Cost, 24h
Last activity
Ticket Triage & Routing
Support · Atlas-M
PN
Priya Nair
Healthy
4,820
98.2%
340ms
$128.40
3 min ago
Invoice Reconciliation
Finance Ops · Atlas-L
MW
Marcus Webb
Warning
1,240
91.4%
1620ms
$412.10
8 min ago
Contract Clause Extraction
Legal · Atlas-L
ET
Elena Torres
Healthy
640
99.1%
890ms
$96.20
14 min ago
Sales Lead Qualification
RevOps · Atlas-M
PN
Priya Nair
Critical
2,140
76.3%
2100ms
$301.80
1 min ago
Customer Churn Risk Summary
Customer Success · Orbit-Turbo
DO
Devon Okafor
Healthy
380
97.6%
1100ms
$54.30
22 min ago
Refund Approval Agent
Finance Ops · Atlas-M
MW
Marcus Webb
Warning
890
94.8%
760ms
$188.90
6 min ago
Knowledge Base Answer Agent
Support · Orbit-Turbo
PN
Priya Nair
Healthy
6,210
96.9%
420ms
$210.50
1 min ago
Vendor Onboarding Doc Review
Procurement · Atlas-L
ET
Elena Torres
Paused
0
—
—
—
2 days ago
Evaluation & Quality
Are our AI outputs good, and are they improving?
All Workflows
Run Evaluation
Lead-Quality-Bench v6
RevOps · Scoring Agent
critical
71.4%
-22.8 pts vs last run
98%
87%
77%
66%
Support-Answer-Accuracy v12
Support · Answer Composer
92.7%
+3.1 pts vs last run
96%
92%
88%
84%
Clause-Extraction-Precision v4
Legal · Clause Classifier
96.5%
+1.7 pts vs last run
100%
96%
93%
90%
Model comparison — Scoring Agent
v2.3 (current) vs v2.4 (candidate)
Metric
v2.3 (live)
v2.4 (candidate)
Accuracy on Lead-Quality-Bench
82.1%
89.4%
Hallucination rate
4.8%
2.1%
Avg latency / run
410ms
690ms
Cost / 1,000 runs
$6.40
$11.20
v2.4 recovers most of the accuracy lost to the ClearbitX outage and cuts hallucinations by more than half at roughly 70% higher cost per run. Recommended for promotion once Enrichment Agent retries are fixed.
Failed examples — Lead-Quality-Bench v6
View workflow →
Input
Expected
Actual
Reason
northfield-labs.io / VP
Engineering
firmographic:
populated
firmographic:
null
enrichment_timeout
brightcove-labs.com /
Head of RevOps
deal_score: 0–100
deal_score:
null
enrichment_timeout
dunemetrics.ai / Founder
qualified: true (ICP
match)
qualified:
false
low-confidence fallback
scoring
harborsuite.io / VP Sales
firmographic:
populated
firmographic:
null
enrichment_timeout
Alerts & Investigation
What needs attention, and what to do next.
All Severities
Open + Investigating
critical
ALT-1082 · Success rate fell below 80% threshold
investigating
Sales Lead Qualification
24h success rate for Sales Lead Qualification is 76.3%, down from a 97% baseline.
Suggested next step — Add retry/backoff to the Enrichment Agent ClearbitX call, or fail runs closed instead of writing null scores.
Today, 08:52
PN
Priya Nair
high
ALT-1079 · P95 latency exceeded 2.0s SLA
monitoring
Invoice Reconciliation
Ledger Matcher step has breached the 2.0s SLA for 3 consecutive hours.
Suggested next step — Vendor ERP API quota was recently lowered — request a rate-limit increase or add request batching.
Yesterday, 21:14
MW
Marcus Webb
medium
ALT-1075 · Cost per run up 34% week-over-week
acknowledged
Refund Approval Agent
Risk Review Agent now issues an extra fraud-signal lookup on 60% of runs.
Suggested next step — Gate the extra lookup behind a $500+ refund threshold instead of running it on every case.
2 days ago
MW
Marcus Webb
low
ALT-1071 · Answer relevance score dipped 3.1 pts
resolved
Knowledge Base Answer Agent
Weekly eval flagged a stale KB index following a documentation migration.
Suggested next step — Resolved — Retriever index was refreshed and relevance recovered to baseline.
3 days ago
PN
Priya Nair
critical
ALT-1068 · Schema validation failures spiked to 18%
resolved
Contract Clause Extraction
Upstream document template changed indentation, causing the parser to mis-split clauses.
Suggested next step — Resolved — Document Segmenter patched to handle the new template; failures back under 1%.
6 days ago
ET
Elena Torres