Insight engine

Your agents, running on your data

Every time a source loads, a roster of agents re-reads the model, finds what changed and why (including things no chart shows), sizes the ₹ impact and routes it to an owner. Every number is computed from your data; nothing is free-text generated.

Agents active

8 / 9

1 blocked by a failing feed

Data loads today

0 rows since 00:00

Insights this week

1 not on any chart

₹ impact surfaced

₹214.0 Cr

this week, de-duplicated by insight

Precision

84%

findings rated useful

How the engine works

One loop on your data: load, model, watch, explain, act, and learn from your feedback.

Sources

14 connected

ERP, MES, HRMS, EHS, banks, Excel, CRM, events

ETL

0 loads today

0 rows · validated & quarantined

Canonical model

23 entities

contracts → lots → cash, one key per thing

Agents

9 agents · 11 findings

detect · explain · recommend

You

0 alerts · 0 actions

inbox, alerts, 06:30 brief, actions

Feedback

84% useful

retunes thresholds and ranking

Your feedback flows back to the agents: 👍/👎 on a finding changes its threshold and how it ranks next time.

Agent roster

· click an agent for what it runs, its findings and its run log

Ingestion & quality agent

Needs attention

After every load and upload · owner CFO

Checks every load: maps new files, runs data-quality rules, quarantines bad rows and tells you what changed in the data.

SAP S/4HANA Finance MES HRMS EHS incident register Stores ledger Compliance registers (Excel)Bank BG statements (5 banks)

Last run

yesterday 06:00

Next run

on next event

This week

0 findings

Precision

95%

Revenue agent

Idle · watching

Daily 06:00 + every CRM stage change · owner CMO

Watches sales, targets and the CRM pipeline for target gaps, customer and region shifts, deal slips and where the quarter will land.

SAP S/4HANA Sales targets (AOP FY27)CRM KPI snapshot store

Last run

yesterday 06:00

Next run

tomorrow 06:00

This week

1 finding · ₹7.3 Cr

Precision

82%

Cash agent

Blocked

Daily 06:00 + bank statement loads · owner CFO

Watches receivables, payables and bank guarantees for DSO creep by customer type, overdue concentration and credit headroom.

Finance SAP S/4HANA Bank BG statements (5 banks)

input down: Bank BG statements (5 banks)

Last run

yesterday 06:00

Next run

tomorrow 06:00

This week

2 findings · ₹86.8 Cr

Precision

79%

Operations agent

Running

Every MES batch (15 min) · owner COO

Watches every MES batch for OEE drift, loss causes, rejects and plant ramp risk.

MES QMS

Last run

yesterday 05:55

Next run

yesterday 06:10

This week

2 findings · ₹5.4 Cr

Precision

80%

Supply-risk agent

Idle · watching

Daily + vendor-portal events · owner COO

Watches POs, vendors and stock for sole-source lateness and traces it to work orders, downtime and revenue at risk.

SAP S/4HANA Stores ledger Event bus

Last run

yesterday 06:00

Next run

on next event

This week

3 findings · ₹18.1 Cr

Precision

83%

People & safety agent

Idle · watching

On HRMS / EHS events + weekly · owner COO

Watches HRMS and EHS for attrition hotspots, leadership exits, staffing ramps and leading safety signals.

HRMS EHS incident register Event bus

Last run

yesterday 05:42

Next run

11 Nov 05:42

This week

0 findings

Precision

73%

Compliance agent

Idle · watching

Daily + on register upload · owner CEO

Watches export licences, IC%, PIL embargoes and offsets, and links each to the shipments and contracts it can block.

Compliance registers (Excel)SAP S/4HANA

Last run

yesterday 06:00

Next run

on next event

This week

3 findings · ₹96.4 Cr

Precision

93%

Root-cause agent

Running

On every Ask why / when another agent hands off · owner CEO

Explains any number on demand: decomposes the change, follows linked signals across tabs and names the drivers.

SAP S/4HANA Finance MES HRMS Stores ledger

Last run

yesterday 06:00

Next run

on next event

This week

0 findings

Precision

84%

Briefing agent

Idle · watching

06:30 daily + before reviews · owner CEO

Ranks everything the other agents found by ₹ impact and owner, and writes each CXO's morning brief.

Event bus KPI snapshot store

Last run

yesterday 06:00

Next run

06:30 today

This week

0 findings

Precision

90%

Live activity

ETL run → agent → insight → notification, newest first

06:30 todayBriefing
Briefing agent ranked 11 findings by ₹ impact and owner; briefs ready for CEO, CFO, COO, CMO
06:02 todayRevenue
AVS-9: 1 milestone late, worst "Delivery tranche 7/11" 36 days (₹7.3 Cr) → notified CEO
Open insight
06:02 todayCash
Canara Bank BG limit 84% used; Canara Bank figures are stale (₹67.1 Cr) → notified CFO
Open insight
06:02 todayOperations
Customer-reported critical NCR NCR0126 open 12 days, RAF-WH (₹3.1 Cr) → notified COO
Open insight
06:02 todaySupply risk
4 single-source vendors below 70% OTIF (90 days). Largest: Aerotitan Alloys (Import) at 46% (₹16.4 Cr) → notified COO
Open insight
06:02 todayCompliance
Export licence SCOMET/2026/4100 (Germany) expires in 45 days (₹71.4 Cr) → notified CMO
Open insight
06:01 todayIngestion
Ingestion & quality agent ran on 8 sources: 0 findings
06:01 todayRevenue
Revenue agent ran on 4 sources: 1 finding
06:01 todayCash
Cash agent ran on 3 sources: 2 findings, 1 not on any card
06:01 todayOperations
Operations agent ran on 2 sources: 2 findings
06:01 todaySupply risk
Supply-risk agent ran on 3 sources: 3 findings
06:01 todayPeople & safety
People & safety agent ran on 3 sources: 0 findings
06:01 todayCompliance
Compliance agent ran on 2 sources: 3 findings

Build your own

Your rules, on any data you have connected

Describe a watcher in plain words, for example “tell me when any sole-source vendor's OTIF stays below 70% for two weeks”. See it as a rule on the canonical model, backtest it on history, then switch it on with an owner and a channel.

Works on any entity, including a file you upload today
Backtest shows how often it would have fired, and whether it was early
Runs after every load, like the agents above

0 watchers active

Open the watcher builder