AI isn’t the problem. AI readiness is.

1. 01

   Ingest your ERP data

   Your ERP extracts and supplier documents come into the PurveO Data Lens read-only, encrypted and held in-region, and each file is detected by type.

2. 02

   Validate the data and build Golden Records

   A quality gate checks coverage and join rates within 48 hours, duplicates merge into one verified Golden Record per supplier, and the Data Readiness Report lists the fixes.

3. 03

   Classify every spend line

   Every spend line is classified by category and channel and mapped to your chart of accounts, so spend reads the same way in every entity.

4. 04

   Measure where you stand

   The AI Readiness Index scores six dimensions from Level 1, Fragmented, to Level 5, AI-native, and KPIs are baselined on your own data.

5. 05

   Codify insights into rules

   Each insight becomes a rule you approve, backtested on 24 months of your history, and goes into Rule Library v1.0.

6. 06

   Apply the rules with agents

   Agents act only inside approved rules and cite the rule behind every action, earning autonomy rule by rule, from L0 Inform to L4 Autonomous.

1. 01

   Mobilise

   Week 0

   Kick-off, data owners named, goals weighted.

   • Signed scope
2. 02

   Ingest and verify

   Weeks 1–2

   Quality gate, Golden Records built, supplier checks run.

   • Data Readiness Report
3. 03

   Diagnose

   Weeks 3–4

   Spend classified, lenses run, SWOT with value at stake.

   • AI Readiness Index baseline
4. 04

   Codify

   Weeks 5–6

   Rules shaped and backtested, approved by you, shadow mode on.

   • Rule Library v1.0
5. 05

   Operate

   Ongoing

   Agents run approved rules; monthly review, quarterly index.

   • Monthly operations report

Overrides and new data feed back into the diagnosis every month. A Supplier Trust-only engagement takes four weeks.

1. 01

   One request letter

   A named owner per dataset, with templates

2. 02

   Five business days

   To upload the seven files

3. 03

   Quality gate in 48 hours

   Coverage and join rates checked

4. 04

   Data Readiness Report

   Scores, gaps and the fixes, per dataset

Encrypted at rest and in transit, held in-region, processed under the UAE and KSA PDPL and India's DPDP Act.

A self-check from your answers. The Data Lens measures the same six dimensions from your actual data.

• Supplier integrity

  How trustworthy is your supplier master?

  • Duplicates, unverified bank and tax data
  • Checked once, at onboarding
  • One clean record, re-checked on a schedule
  • Verified continuously, by risk

• Spend visibility

  How well can you see what you spend?

  • Only by GL account
  • Some categories, in spreadsheets
  • Most spend classified by category
  • All spend classified, near real time

• Contract coverage

  Where do your contracts live?

  • No central register
  • A partial spreadsheet
  • A complete register with key terms
  • Terms enforced on every transaction

• Process discipline

  Does a PO come before the invoice?

  • Rarely: non-PO buying is common
  • For some categories
  • Yes, as the norm
  • Always: only exceptions need a person

• Policy codification

  How is your buying policy enforced?

  • It lives in PDFs and people's heads
  • Written limits, checked by hand
  • Approval limits enforced in the system
  • A full rule library, tested and monitored

• Data access

  How do you get data out of your ERP?

  • Manual extracts, inconsistent IDs
  • Regular exports, with gaps
  • Scheduled extracts that pass a quality check
  • Live ERP integration

• AI-ready

  Your foundations are strong. Agents can take on proven rules now, inside the limits you set.

• Nearly AI-ready

  Most foundations are in place. Agents can start on approval while the Data Lens closes the last gaps.

• Ready to start

  The groundwork is partly there. Six weeks of supplier verification and rules make it ready for agents.

• AI needed

  Manual work is holding your procurement back. AI is how you fix it: verify suppliers, classify spend and turn policy into rules.

What is AI readiness in procurement?

AI readiness in procurement means your suppliers are verified, your spend is classified and your policy is written as rules a system can enforce. Only then can AI agents act safely. PurveO measures it with the AI Readiness Index: six dimensions, each scored from Level 1, Fragmented, to Level 5, AI-native.

How long does it take?

Six weeks from raw extracts to an approved Rule Library: mobilise in week 0, ingest and verify in weeks 1–2, diagnose in weeks 3–4 and codify in weeks 5–6. A Supplier Trust-only engagement takes four weeks.

What data do you need to start?

Seven datasets every mid-market ERP holds: the supplier master, 24 months of AP invoice lines and payments, the chart of accounts, cost centres, your procurement policy and approval limits, and the supplier documents on file. Each has a fallback if it is missing.

What do we receive?

A Data Readiness Report, a verified supplier Golden Record and status list, a classified spend cube, an insight and SWOT pack with value at stake, your AI Readiness Index baseline, an approved Rule Library and a 90-day roadmap.

Does it change anything in our ERP?

Not to start: the work begins from read-only extracts. Only changes you approve, such as the cleaned supplier record, are written back, and your ERP stays the system of record.

Who is it built for?

Companies with USD 25–250M revenue, 10–50 procurement users and 300 or more requesters, often across several entities, running NetSuite or Business Central first.

How is our data protected?

Bank and ownership data is encrypted at rest and in transit, held in-region with UAE cloud residency, and processed under the UAE and KSA PDPL and India's DPDP Act. Access is role-based and every view is logged.