The challenge
Demand signals across the dealer network arrived too late to flex plant production cadence, and procurement RFQs for line-side components ran on long email chains with no consolidated history. Holding-cost and stock-out exposure both ran higher than the operations team wanted, especially on monsoon-season SKUs.
How we approached it
Ajuni put Nila on a 48-hour rolling forecast tied to SAP S/4HANA and dealer DMS signals, and wired Dwight into the procurement workflow for RFQ orchestration against vendor history. The deployment inherited Hero's existing Azure India tenant, with a Snowflake mirror for analytics replay.
Outcomes in production
- 92% accuracy on 48-hour rolling demand forecast
- RFQ-to-PO cycle compressed to 4 days
- ₹4.2 cr / year in working-capital savings
- Stock-out exposure cut on monsoon-season SKUs
- 10 weeks kickoff to plant-floor production
Stack & guardrails
Integration & deployment
- Azure India · Hero tenant
- SAP S/4HANA integration
- Dealer DMS connector
- Claude Sonnet 4 · API
- Snowflake mirror · analytics
- Edge inference · plants
Compliance & audit
- IATF 16949
- ISO 27001
- DPDP Act 2023
- Internal SOX review
Timeline
-
Week 0–2
Forecast scoping
Demand signal sources ranked with operations; 48-hour window confirmed.
-
Week 3–5
ERP + DMS wiring
Nila integrated with SAP and dealer DMS; forecast accuracy validated.
-
Week 6–8
Shadow procurement
Dwight drafted RFQs alongside buyers; vendor history surfaced inline.
-
Week 9–10
Plant rollout
Forecast and RFQ agents live; working-capital impact reviewed monthly.