No AI strategy tied to the P&L
Ambition lives in slides. There is no portfolio, no sequencing, no owner per outcome.
Services/01 For enterprises
AI Program
A program-level engagement for organisations that want AI to move the company, not just a department. We run strategy, delivery and adoption in parallel with your leadership, governance and operating teams.
This is what a company without an AI plan sounds like.
These questions now land with whoever happens to sit nearby: a team lead, IT, the exec team. Everyone gives their own answer and next week they are back on the table. Meanwhile:
- Prototypes never reach productionengineers keep building, but there is no foundation to ship on
- Spend grows without a view on impactinvoices go up, the business case is missing
- Legal blocks, teams waitevery use case turns into a discussion of weeks
- Adoption fadeslicences keep running, usage drops off after week two
The AI Program tackles this as one whole: strategy, governance, foundation and adoption, run in parallel, with an owner per outcome.
- #directie · Joost · COO: We now have three departments each running their own AI pilot. Who decides which one we scale?
- #engineering · Sanne · Lead Engineer: Every team is using a different model on its own key. Do we want to centralise this or leave it as is?
- #finance · Mark · FP&A: This month's OpenAI invoice is more than 3x the budget. Can someone walk me through where that came from?
- #legal · Eva · Legal Counsel: I'm seeing people paste customer names into ChatGPT. We have no policy and no approved tool. What do we do?
- #people · Amira · People Lead: Two candidates this week asked which AI tools they can use here. I didn't have a good answer.
- #platform · Tim · Platform Engineer: The support agent has been live since last week. I have no way to see if the answers are getting better or worse.
- #mt-weekly · Hanna · CEO: I've had four requests for an 'AI project' land on my desk this week. How do we weigh these against the roadmap?
- DM · Mark → Joost · Mark · FP&A: The board wants an ROI figure on AI next month. I have the costs, but nothing measurable to put against them.
- #directie · Joost · COO: Legal wants a four-week review on every use case, product wants to go live next week. How do we solve this structurally?
- #people · Amira · People Lead: Copilot licences have been running for 6 weeks. Active use is below 20%. Push through or scale back?
- #engineering · Noor · Developer: I've got a prototype that handles 40% of my tickets. How do we get this off my laptop and into production?
- #directie · Hanna · CEO: Everyone shows up with a list of processes that are 'perfect for AI'. What do we actually base the choice on?
The problem
AI is seen as interesting, but the plan to roll it out at scale is missing.
Most enterprises now use AI in pockets. What is missing is a broader program: an AI strategy tied to the P&L, a data and platform foundation, structured use case exploration, and adoption inside the teams that do the work.
Foundation is not ready
Data, identity and evaluation are fragmented. Agents cannot reach context or systems safely.
Use case exploration is ad hoc
Pilots are picked by enthusiasm, not by impact. Nothing gets sequenced towards production.
Adoption is left to chance
Tools are rolled out, not embedded. Teams keep working around the AI instead of with it.
What you get
Concrete deliverables.
AI strategy and roadmap
Use case portfolio scored on impact, feasibility and risk, sequenced over 3 to 6 quarters and tied to P&L targets.
Governance and operating rhythm
Steering committee, intake board, evaluation rituals and a model for prioritising squads against shifting business priorities.
Production agents in core processes
Working systems in commercial, operations, support or finance - shipped, measured and handed over.
Adoption and enablement
Champion training, change management, internal documentation and a capability uplift programme for your teams.
Data and platform foundation
The connectivity, identity and evaluation layer your agents need to run safely in production.
Reporting on business impact
A quarterly view of cost, throughput, revenue and risk movement directly attributable to the program.
How we deliver
From ambition to agents in production.
01
Session
Kickoff
Executive working session to align ambition, P&L targets and the first wave of use cases.
02
Scan
Discovery
Six to ten weeks of structured assessment across business units. Output: prioritised portfolio plus operating model.
03
Build
Delivery
Parallel squads ship agents into the highest-value processes, in rolling 6 to 12 week cycles.
04
Squad
Ongoing
Embedded senior engineers and PM continue to ship and maintain, billed monthly.
Timeline
3 to 12 months, rolling
Investment
From €75k per quarter
Fit
Who this is for. And who it is not.
Fits you if
Not the fit if
- 01250+ employees with multiple business units or geographies
- 02Leadership ready to commit a steering committee and a budget
- 03Existing data, CRM and ERP stack to integrate with
- 04Targets at the P&L level - revenue, margin, cost-to-serve
- 01You want a single bounded pilot - look at AI Product Development
- 02You are a founder-led team under 50 - the Centralized Database Model fits better
- 03You are scouting vendors with no executive mandate yet
Agentic systems we build.
Production-grade systems we design, build and deploy for commercial teams. Each one tailored to a single high-value workflow.
Your directAI partners.
01Derk Disselhoff
Founder & Director
02Elko Lemiso
AI Engineer