Buy best
Worked example
Customer service / contact centre AI
Practice 01 · AI Strategy, Roadmap & Procurement
Decide what to build, what to buy, and in what order - defensibly.
Bouddi works alongside business owners and technology teams to discover how the work actually runs, define an AI investment thesis, sequence the roadmap, and procure the platforms and vendors that will sit underneath it - with the same evidence discipline our governance practice is known for.
3
Engagement pillars - Strategy, Roadmap, Procurement - modular or combined
8–14
Weeks for a typical end-to-end engagement, depending on scope
2-side
Co-delivered with business owners and technology delivery teams
D1–D6
Anchored to the same maturity domains as our governance practice
Why this engagement exists
Most AI investment decisions are made before anyone has done the work to defend them.
A vendor demo lands well. A board asks for a strategy. A technology team is told to "do AI." The result is a portfolio of pilots that don't ladder to a thesis, contracts that don't survive a risk review, and a roadmap that no-one can sequence. Bouddi closes that gap before the spend lands.
What we observe
Across ANZ engagements we see the same five patterns - not because organisations are careless, but because the operating cadence rewards motion over discipline.
This engagement is built to interrupt the pattern: surface the use cases worth doing, disqualify the ones that aren't, and put the procurement and roadmap evidence in place before the cheque is signed.
-
01
Strategy decided in slide form, not investment formAn "AI strategy" exists, but the use cases haven't been costed, sequenced, or tied to measurable business outcomes - so executive committee can't actually fund it.
-
02
Vendor selected before the requirement was writtenProcurement is reverse-engineered to justify a platform the business has already chosen, leaving the technology team with integration debt and the risk team with contracts they can't defend.
-
03
Roadmaps without dependency or capacity logicUse-case lists masquerading as roadmaps - no sequencing against data readiness, MLOps capability, talent, change capacity or regulatory constraints.
-
04
Build vs. buy decided by who shouted lastNo total-cost-of-ownership view, no honest assessment of internal capability, and no clear pattern for which workloads belong on a vendor stack vs. an internal platform.
-
05
Governance bolted on after the factStrategy and procurement decisions get made first; risk, audit and assurance arrive late and end up reverse-engineering policy to fit choices already made.
The foundation
We map the work before we map the AI.
Most AI strategy fails at the same place: a use case looks compelling on a slide, but the team doing the work knows it solves a problem that doesn't exist, or skips the constraint that actually matters. Customer business process discovery is the foundation underneath every Bouddi strategy, roadmap and procurement engagement.
What process discovery does
We start with what you already have. Existing process maps, BPMN models, operating procedures, value-stream analyses, six-sigma artefacts, customer-journey work - we inherit, validate and extend rather than rebuild. Where the documentation doesn't exist, we sit with your operators, customers and technology teams and capture how the processes actually run today - including the workarounds, the manual hand-offs, the data that lives in spreadsheets, and the decisions that depend on someone's experience.
The result is a process baseline that does three jobs at once: it surfaces the AI use cases worth pursuing and disqualifies the ones that aren't; it gives the technology team the requirement clarity they need to architect, integrate and procure; and it gives risk, audit and finance a defensible record of where AI will sit inside the operating model - before, not after, the spend.
-
D1
Customer journey & value-stream mappingEnd-to-end customer journeys and internal value streams documented from the operator's perspective, not the org chart's. Surfaces the moments that matter and the seams where work breaks.
-
D2
Process inventory & annotationEvery in-scope process catalogued with frequency, volume, system, owner, decision rights, exception rate and risk profile - so prioritisation has data behind it.
-
D3
Pain-points, constraints and workaroundsWhere the work breaks, what people do to keep it moving, and which constraints are real (regulatory, contractual, customer) versus inherited.
-
D4
AI & automation opportunity registerProcess steps tagged for AI, automation, redesign or "leave alone", with effort, value and confidence annotations - the direct input to the use-case portfolio.
-
D5
Future-state process designFor tier-1 candidates, a target-state process model showing the human / AI / system division of labour, decision points, controls and measurement - ready to feed roadmap and procurement.
Method: working sessions with operators and front-line teams, customer journey workshops, document and system-trace review, BPMN-aligned notation. Cross-validated against customer outcomes, not internal narrative.
What we do
Three pillars. Run modularly, or as one connected engagement.
Each pillar is a defensible deliverable on its own. Run together, they form an end-to-end decision pack - the artefacts your executive committee, technology leadership, finance and risk functions all need to align before money or signatures move.
Pillar 01
AI Strategy
The investment thesis. Why we are doing AI, where the value is, what we will not do, and how success is measured. Written to be tabled at executive committee or board.
Outputs include
- AI ambition statement and value framework
- Use-case discovery and qualification
- Tier-1 use cases costed and prioritised
- Build / buy / partner disposition per use case
- Target operating model for AI delivery
- Investment case and KPI scaffold
Pillar 02
AI Roadmap
The sequenced delivery plan. What gets built when, by whom, on what platform, with what dependencies - paced against your real data, MLOps and change capacity.
Outputs include
- AI delivery capability assessment (data, platform, MLOps, talent)
- 12–24 month sequenced roadmap with dependencies
- Reference architecture and platform pattern
- Capacity, resourcing and skills plan
- Investment phasing and stage-gate criteria
- Roadmap briefing pack for executive and board
Pillar 03
AI Procurement
The evidence underneath the cheque. Vendor evaluation, RFP support, contract review and total-cost modelling - done so the choice survives risk, finance and audit review.
Outputs include
- Requirement specification and evaluation criteria
- Vendor landscape and shortlist
- RFP / RFI authoring and response evaluation
- Total cost of ownership and build-vs-buy model
- AI-specific contract clause library
- Vendor due diligence and MSP assessment
Strategy Lens · Stage 02 of the Bouddi Framing Process
The Bouddi Investment Position Matrix.
Every Strategy engagement turns on a single question: of all the AI bets we could place, which ones should we place, in what shape, in what order? The Investment Position Matrix is the analytical lens that answers it. Two axes - Strategic Centrality (how core to differentiation: alignment, advantage potential, value-chain integration) and Defensibility (how unique we can make our position: data uniqueness, IP potential, switching costs, vendor maturity, rate of commoditisation) - produce four clear positions for any candidate AI investment.
Low Defensibility
High Defensibility
High Strategic Centrality
Build
Worked example
Proprietary credit risk modelling
Low Strategic Centrality
Wait
Worked example
General-purpose AI productivity tools
Partner
Worked example
Specialist RegTech (AML / sanctions screening)
Each position in detail.
The verb is the headline. The capital allocation, the timeline, and the conversation with the board are what actually changes.
Build
Proprietary credit risk modelling
High Strategic Centrality · High Defensibility
Lending is the core profit engine and the customer transaction data is uniquely yours. A vendor model gives away your most differentiated asset and produces something a competitor can replicate. Worth the long timeline and the investment in internal model risk capability - this is where engineering capital deserves to land.
Buy best
Customer service / contact centre AI
High Strategic Centrality · Low Defensibility
Customer service shapes the brand - getting it wrong drives churn. But intent recognition, agent assist, knowledge retrieval and automated routing are commoditised; Microsoft, Salesforce, ServiceNow and a dozen specialists ship comparable platforms. Building this internally would burn 18–24 months and tens of millions to land where you could be in 90 days. Buy the leader and redirect the engineering capital somewhere defensible.
Partner
Specialist RegTech (AML / sanctions screening)
Low Strategic Centrality · High Defensibility
AML transaction monitoring is mandatory but not differentiating - no bank wins customers on the quality of its AML system. Yet the capability is genuinely defensible (deep regulatory expertise, specialised data, complex models). Partnering with a specialist (or acquiring one) gets the defensibility without the strategic distraction. Building consumes risk and compliance engineering capacity better spent elsewhere; buying off-the-shelf gives you weak capability.
Wait
General-purpose AI productivity tools
Low Strategic Centrality · Low Defensibility
Every vendor in 2026 is pitching enterprise-wide rollouts of meeting AI, agent assistants and "AI everything." None are strategically central - they're cost-of-doing-business at best. The capabilities are commoditising fast: prices are falling, free options are emerging, consolidation will leave two or three winners. Wait 12–18 months, watch the market consolidate, then buy the obvious winner cheaply. The cost of waiting is small; the cost of betting wrong is real.
Self-serve register
Investment Position Matrix canvas
A downloadable canvas your executive team can place candidate AI investments onto in an afternoon. Useful for board pre-reads, investment committee workshops, and the working sketch you bring into a Strategy engagement.
Engaged register
Calibrated, evidence-validated placement
Delivered inside a Strategy engagement: each placement is justified against your strategic intent, your data position, your competitive landscape and the rate at which the capability is commoditising in your market - and signed off by the relevant business sponsor before it goes to the board.
Who we work with
Co-delivered with the business and the technology team in the same room.
Strategy without delivery is theatre; delivery without strategy is debt. Every engagement is run with a named business sponsor and a named technology lead, with both sides accountable for the artefacts that come out.
Customer side · Business owners
For the executives funding the work.
CEOs, COOs, CFOs and line-of-business leaders who need to know what AI is worth to their P&L, where to place the bets, and how to defend the investment to their board, customers and regulators.
Typical sponsors: CEO, COO, CFO, Chief Strategy Officer, Heads of Business Lines,
Chief Customer Officer, Audit Committee Chair.
Technology side · Delivery leaders
For the teams that have to make it real.
CTOs, CIOs, Heads of AI and Chief Data Officers who need a roadmap their team can actually deliver, an architecture that doesn't lock the organisation in, and procurement decisions their engineers will inherit cleanly.
Typical co-leads: CTO, CIO, Chief Data Officer, Head of AI / ML, Head of
Architecture, Head of Engineering, Procurement / Vendor Management lead.
How an engagement runs · The Bouddi Strategy Method
Five phases. Designed to deliver decisions, not decks.
A full Strategy + Roadmap + Procurement engagement runs eight to fourteen weeks depending on scope and the number of vendors in play. Each phase produces a signed-off artefact your team owns, not a Bouddi deck. The full methodology - phases, deliverables per phase, named audiences and bridges to other practices - lives at The Bouddi Strategy Method.
Phase 01
Frame & Align
1–2 weeks
Stakeholder map, success criteria, current AI footprint, scope agreement and decision register set up.
Phase 02
Discover & Diagnose
2–4 weeks
Customer business process discovery, value-stream mapping, use-case qualification, data and capability baseline, build-vs-buy disposition.
Phase 03
Strategy & Roadmap
2–4 weeks
AI ambition, value framework, target operating model, sequenced 12–24 month roadmap, investment case.
Phase 04
Procure
3–5 weeks
Requirements, RFP authoring, vendor evaluation, TCO model, AI contract clauses, due-diligence pack.
Phase 05
Mobilise
1–2 weeks
Programme structure, decision rights, executive briefing pack, hand-off to delivery and to governance.
Sample deliverables
A representative slice - full deliverables list is shaped to engagement scope.
D.1
Customer Business Process Map
Current-state customer journeys and value streams documented from the operator's perspective, with frequency, volume, system, owner and exception annotations.
bpmn · pdf · Discovery
D.2
AI & Automation Opportunity Register
Process steps tagged for AI, automation, redesign or "leave alone", with effort, value and confidence - the direct input to the use-case portfolio.
xlsx · Discovery
S.1
AI Ambition & Value Framework
Where AI matters most to the P&L, where it doesn't, and how value will be measured. One signed-off page plus working detail.
docx · pptx · Strategy
S.2
Use-Case Portfolio
Long-listed, qualified and tier-1 prioritised use cases with effort, value, risk and dependency annotations.
xlsx · Strategy + Roadmap
S.3
Build / Buy / Partner Disposition
Decision-grade rationale for each tier-1 workload - internal build, vendor platform, partner co-build, or wait.
docx · xlsx · Strategy
R.1
12–24 Month Roadmap
Sequenced delivery plan with owner, dependency, stage-gate, capacity and outcome on every initiative.
xlsx · pptx · Roadmap
R.2
Reference Architecture & Platform Pattern
A target architecture pattern that integrates with existing data, identity, observability and risk control planes.
svg · pdf · Roadmap
P.1
Vendor Shortlist & Evaluation Pack
Weighted scoring against documented requirements, with technical, commercial, risk and fit dimensions.
xlsx · docx · Procurement
P.2
TCO & Build-vs-Buy Model
Three- and five-year total cost of ownership with build, buy and hybrid scenarios, plus exit-cost estimate.
xlsx · Procurement
Want to see what these deliverables actually look like?
Sample artefacts - customer process map, vendor evaluation pack, 12-month roadmap, AI contract clauses - are available in our prospect portal.
Open the prospect portal →
Engagement options
Modular by design. Three ways in.
Pick a single pillar where the heat is highest, or run the whole pathway. Indicative fees are scoped against your environment before a fixed price is offered. Indicative pricing is available in the Prospective Clients area.
Already running governance work with us?
Strategy, Roadmap and Procurement engagements run cleanly alongside the AI Governance
Diagnostic and Implementation services - same methodology, same evidence discipline,
artefacts cross-reference rather than duplicate.
Start here
A confidential conversation, not a pitch.
Tell us where the heat is - a vendor decision under time pressure, a roadmap your board isn't buying, a strategy that needs sharpening. The first thirty minutes are no-obligation and covered by NDA on request.