How we engage
The Ubuntu Method.
Strategy through implementation, in one engagement. Four phases that take an enterprise AI initiative from idea to production in weeks, not quarters.
Most clients start with Discover. Each phase is scoped and priced separately.
- 012-3 weeks
Discover
We map your existing tech stack, your data flows, and the business problems where AI can actually move the needle. We talk to your engineers, your operators, and your business owners. We read your code, your schemas, and your runbooks. We come out with a prioritized opportunity map, ranked by feasibility and impact.
What we ship
- ▪Current-state architecture map
- ▪Data inventory and flow diagrams
- ▪Prioritized opportunity register (5-12 candidate projects)
- ▪Recommendation: what to build first, second, never
- 021-2 weeks
Strategize
We design the AI system architecture for the highest-priority opportunity. What to build, what to buy, what to integrate, what to defer. Model selection, infrastructure sizing, security posture, evaluation strategy. Outcome: a build plan with measurable success criteria.
What we ship
- ▪Solution architecture diagram
- ▪Model selection rationale
- ▪Infrastructure and GPU sizing plan
- ▪Evaluation suite design
- ▪Success criteria with quantitative targets
- ▪Build plan with milestones
- 034-8 weeks
Build
Our engineers implement on your infrastructure. Daily standups, weekly demos to your stakeholders, biweekly delivery of working software. You own the code from the first commit, the model weights from the first training run, and the pipelines from the first deployment.
What we ship
- ▪Production AI system, deployed
- ▪Source code in your repository
- ▪Model weights in your storage
- ▪Eval pipeline running on every model update
- ▪Monitoring dashboards in your observability stack
- ▪Runbooks for common operations
- 04Ongoing or 1-2 week handover
Operate
We hand over with documentation, training, and an optional ongoing support engagement. Your team learns the system from us, not just how to operate it but why it's built the way it is. When we leave, your team runs it. If you want, we stay on call.
What we ship
- ▪Architecture documentation
- ▪Operations runbooks
- ▪Training sessions for your team
- ▪Optional ongoing retainer for incidents and model updates
Common questions
How engagements work
- How long does a typical engagement take?
- First production AI system in 6-12 weeks. The Discover phase is typically 2-3 weeks, Strategize 1-2 weeks, Build 4-8 weeks, and Operate handover 1-2 weeks.
- Do you take equity, fixed fee, or time-and-materials?
- Fixed fee per phase by default. Time-and-materials is available for ongoing Operate engagements. We do not take equity in client companies.
- Who owns the code and the models when the engagement ends?
- You do. The code, the model weights, the training pipelines, and the deployment runbooks are yours. We hand over with documentation and training.
- Do you sign NDAs and BAAs?
- Yes. We routinely sign NDAs, BAAs (for healthcare in the US), and equivalent data-protection agreements for our other regulated industries.
The closer
Build the AI you'd be proud to own.
Thirty minutes to talk through your stack, your data, and the AI opportunity you care about most. No pitch deck. No sales theatre.