Case Study: How AI-Driven Vendor Management Transformed a Housing Society
By Manish Banda — 2026-09-04
This illustrative case study shows how a housing society improved vendor management, and is honest about where AI helped and where it didn't.
In short: this illustrative case study shows how a housing society improved vendor management, and is honest about where AI helped and where it didn't. Most of the gains came from structured records: a vendor register, competing quotes, purchase orders, work orders and approvals. AI helped around the edges, mainly with communication.
What role can AI play in vendor management?
A modest one. Vendor management depends on good records and fair processes: who the vendors are, what they quoted, what was ordered, what was delivered and what was paid. AI can help draft clear notices about vendor work, help residents report problems with a vendor's work clearly and explain spending in a plain-language summary of the accounts. It doesn't choose vendors or approve payments.
Case study overview: Society X
Society X is an illustrative 350-flat society with around 25 regular vendors: security, housekeeping, lifts, pumps, gardening, pest control and more. Before the change:
- Vendor details lived in the manager's phone.
- Quotes arrived on WhatsApp and were hard to compare.
- Payments were made without clear records of what was ordered.
- Annual maintenance contracts lapsed unnoticed.
- TDS was handled inconsistently.
What did the society implement?
- Vendor register: every vendor with contact, GST, PAN, bank details and payee category.
- Quotation requests: several quotes recorded and compared before choosing.
- Purchase orders: what was ordered, with GST, tracked to delivery.
- Work orders: jobs linked to complaints or assets, with quoted and actual cost.
- Preventive maintenance: servicing raised before due dates, with the contracted vendor.
- Expense approvals: payments approved with bills attached, and TDS tracked.
- AI help: notices about vendor visits drafted quickly, and a plain-language spending summary for the AGM.
See myKutir's society management ERP.
Results observed
- Major purchases always had competing quotes on record.
- Servicing happened on schedule instead of after breakdowns.
- Every payment traced back to an order or work order.
- TDS was handled consistently.
- Members understood vendor spending better at the AGM.
Common pitfalls
- Calling a process "AI-driven" when the real work is good records.
- Keeping quotes in chat, outside the system.
- Paying without a purchase order or work order.
- Letting service contracts lapse without reminders.
Traditional or structured vendor management?
| Area | Traditional | Structured, with AI help |
|---|---|---|
| Vendor details | Manager's phone | Vendor register |
| Quotes | Chat messages | Compared on record |
| Orders | Verbal | Purchase orders |
| Communication | Ad hoc | AI-drafted notices |
Steps for other societies
- Build the vendor register.
- Require competing quotes above an agreed amount.
- Use purchase orders and work orders for all jobs.
- Set preventive maintenance schedules.
- Approve payments with bills attached.
- Use AI for clear communication about vendor work.
Conclusion
Good vendor management is built on records and fair process. AI adds convenience, not the foundation. See also the full guide to vendor management.
Key takeaways
- Most gains come from a vendor register, quotes, orders and approvals.
- AI helps with communication and summaries, not vendor decisions.
- Schedule servicing to avoid breakdowns.
- Trace every payment to an order.