Book a discovery call

17 Kimia Biosciences · India Live, phase one

Kimia Biosciences: purchase requests out of chat and PDFs, onto fixed approvals

One controlled flow on the firm's own server, every decision recorded with its reason.

Before
Purchase requests raised over chat messages and PDFs, approvals nowhere to find later
After
Every request moves through fixed review gates in one system on the firm's own server, each decision recorded with a reason
12 of 12 QA cases passed on camera before handover
The work, before and after
Before and afterBefore: Request in chat, then PDF passed around, then Approved in conversation, then No record. After: Request raised, then Fixed review gates, then Reason recorded, then Decision on record. Result: 12 of 12, QA cases passed on camera before handover.BEFORE4 OF 4 PAIN POINTSAFTER3 OF 4 AUTOMATEDRequest in chatPDF passed aroundApproved inconversationNo recordRequest raisedFixed review gatesReason recordedDecision on record12 of 12QA cases passed on camera before handoverBefore and afterBefore: Request in chat, then PDF passed around, then Approved in conversation, then No record. After: Request raised, then Fixed review gates, then Reason recorded, then Decision on record. Result: 12 of 12, QA cases passed on camera before handover.BEFORE4 OF 4 PAIN POINTSAFTER3 OF 4 AUTOMATEDRequest in chatPDF passed aroundApproved in conversationNo recordRequest raisedFixed review gatesReason recordedDecision on record12 of 12QA cases passed on camera before handover
Every
decision recorded with its reason
Own server
no third-party cloud
How the system runs
Input
  • Purchase request
ArrAI
Review gates
Output
  • Approved, on record
  • Rejected, with reason

01 What changed

  • Approvals are on record instead of in a chat
  • The floor raises requests the same way every time
  • Phase one built and running, with phase two scoped
The full story The problem, how we approached it, and what we built

Kimia Biosciences, a WHO-GMP pharmaceutical manufacturer in India, raised and approved purchase requests across paper forms and chat messages. We built the system that takes a request from raised to approved through fixed review gates, with every decision and its reason on record.

Built for
Purchase team and leadership
Engagement
Kimia Biosciences
Sector
Pharmaceutical manufacturing
Region
India
Year
2026

02 The problem

A purchase request started in a chat, moved as a PDF, and got approved in a conversation nobody could find again. Audits meant reconstructing what happened.

The firm needed control without slowing the floor down, and a record it could show.

03 The approach

We mapped the request path with the people who raise and approve, then built one system that carries a request through its review gates and records each decision with a reason. It runs on the firm's own server, with no third-party cloud.

Phase one is built and running on the firm's beta. Process detail stays inside the firm; this page describes the shape, not the rules.

04 What we built

  • Purchase request intake replacing chat and PDF
  • Fixed review gates with a recorded decision and reason at each
  • A record per request from raised to approved
  • Two people acting on the same request at once cannot push it past a gate

Want this kind of result mapped for your business?

Book a discovery call