Transformation Workshop
From pain points to a funded AI roadmap.
Replace weeks of discovery meetings with one structured, 90-minute session.
Your leaders map the work that slows you down, with a guest CFO in the room to validate it peer to peer. We turn it into scored, prioritised AI use cases, each with its return, productivity, efficiency and risk case, then a fixed-price roadmap within 14 days.
AP invoice validation and coding
Illustrative example, based on patterns from Octane discovery sessions across finance, HR and operations.
Where the workshop fits
From workshop to production, at a fixed cost.
Transformation Workshop
Map the process, score the ideas and agree what success looks like.
Roadmap and proposal
A fixed-price scope and timeline for your top one or two use cases.
Early Access Pilot
Prove it on your own use case, against your criteria. Learn more →
Production
Take it live, then extend to the next process with the same playbook.
Why start with a workshop
Almost everyone uses AI. Few can show it pays.
Of organisations already use AI in at least one business function.
McKinseyOf CFOs say AI has contributed positively to EBIT.
McKinseySaw no tangible enterprise-level EBIT impact from generative AI.
McKinsey survey, 2024The fix, McKinsey argues, is to set measurable targets first, redesign processes end to end and decide where people stay in charge. That's exactly what the 90 minutes are built to do.
Source: McKinsey & Company, How CFOs can make AI pay off. Figures as reported by McKinsey.
Peer-to-peer, not vendor-to-buyer
A guest CFO in the room.
Pain points surface faster when the person asking has sat in your chair.
Every Transformation Workshop includes an experienced guest CFO. They're there for your leadership team, not to sell: to test what you're seeing, compare it with what they've lived through, and help find the pain points that really cost you.
Peer-to-peer validation
A CFO-to-CFO conversation about what's really slowing finance down, and whether it's worth fixing first.
The questions only a CFO asks
Where does the close stall? Which controls keep your auditors up at night? What would you do with the time back?
What worked, and what didn't
Lessons from their own transformation, so your roadmap avoids the traps and starts where value is proven.
An honest view of the business case
A second finance leader pressure-tests the value, costs and risks before anything reaches your board.
The AI value chain
From the pain you feel to value you can measure.
Agents add the most where work is high-volume, repetitive, data-heavy or spread across systems, with people stepping in on the complex cases. Every use case we shortlist is sized on four measures: return, productivity, efficiency and risk.
Supplier overbilling targeted for recovery, against ~10% captured by hand
McKinsey, utility exampleA month freed in accounts payable at 10,000 invoices a month
Octane discovery estimateSubledger upload and reconciliation in a month-end close build
Octane client buildAccuracy across 2,000+ KPIs in an agent-run performance review
McKinsey, European bank exampleMonth-end close automation
Supplier overbilling recovery
Rolling forecast on Planning Analytics
Balance sheet and asset reconciliation
Performance review and reporting
HR self-service
Contract review and NDAs
Sources: Octane discovery sessions (anonymised) and client builds; McKinsey & Company, How CFOs can make AI pay off; IBM, AskHR case study; Thomson Reuters, as cited in IBM, AI agent use cases. AP estimate assumes 80% of invoices post touchlessly and about 170 working hours per FTE per month. Benchmarks show what comparable organisations achieved; your own targets are set in the workshop.
How the 90 minutes work
Five steps. One session. A ranked shortlist.
A structured deep dive with your leaders and a guest CFO. Every idea is captured, scored and turned into a business case, so you leave knowing exactly where to start.
Capture every idea, from every team
Guided rounds pull AI ideas from each department, each tied to a real pain point.
See instantly where to start
Teams rate impact and effort, and every idea lands on a priority matrix.
Hover over or tap a bright dot to see a use case. Faint dots are the rest of the 102.
AI scores every use case
Five dimensions, scored out of 100, with an executive summary.
Executive summary: high volume, clear rules and a measurable baseline. Sits on the existing ERP, so no system change. Start here.
A business case, drafted in the room
Problem, solution, agent design and value, ready to cost.
The top 3, ranked for you
Every use case is scored, and the strongest rise to the top.
Ranked from 72 scored use cases, each with an owner and a next step.
One card that keeps everyone focused
Each priority initiative leaves the room defined on a single card.
- Who
- Accounts payable makers and checkers across 700 branches
- Pain
- 10–11 minutes of manual lookups per invoice; ~5% rejected late
- Opportunity
- Agents validate and code every invoice; people handle exceptions only
- Outcome
- Touchless posting for clean invoices; ~1,400 hours a month freed
- Owner
- Head of Accounts Payable, with the CFO as sponsor
- Next step
- Architecture deep dive, then an Early Access Pilot
Where value is made or lost
We walk the journey step by step and mark every decision a person makes today.
| Step | Decision? | Today | With agents |
|---|---|---|---|
| Receive invoice | — | Email and portal, many formats | Captured automatically, any format |
| Extract fields | — | OCR about 85% accurate | Every field quoted with a confidence |
| Validate vendor, GST, TDS | ✓ | Manual system lookups | Checked by agent, evidence cited |
| Assign GL code | ✓ | Maker's judgement | Agent recommends, rules confirm |
| Duplicate and fraud check | ✓ | Caught late, if at all | Blocked before posting |
| Approve by limit | ✓ | Checker reviews every invoice | Routed by rule; exceptions only |
| Post to ERP | — | Keyed by hand | Posted after controls re-checked |
Come prepared
Bring real journeys. Leave with real numbers.
It's a working session. The more of your own reality you bring, the sharper the shortlist and the business case.
What to bring for two or three processes
- A simple process map, with the points where someone makes a decision
- The numbers: volumes, cost, % handled manually, error or exception rates, cycle times
- Known pain points: bottlenecks, rework, delays and sources of risk
- Your ambition: what better looks like, such as a faster close or fewer errors
- Data readiness: where the data lives and any quality gaps
Order-of-magnitude figures are fine. Precision comes later.
Where to look: six signs of value
Leakage we look for
Fraud and losses, revenue leakage, manual review and rework, error and rejection rates, compliance breaches, delays.
Uplift we size
Straight-through processing, cycle time, accuracy, cost savings, productivity, decision consistency.
Bring your hardest process.
90 minutes, with a guest CFO in the room for peer-to-peer validation.
- 3–5 prioritised, high-impact AI initiatives
- A value case for each: return, productivity, efficiency and risk
- A near, mid and long-term roadmap
- Executive alignment, with sponsors and owners named
- A clear go-forward decision, and a fixed-price proposal within 14 days