Should You Build, Hire or Partner for AI Deployment?
Answer twelve questions about your team, the deployment and its constraints. You get a recommended engagement model with the reasoning behind it, not just a score.
No email required to see your result. Vendor-neutral: featured placements are labelled and never affect recommendations.
What it asks
Engineering capacity, AI experience, who the deployment is for, integrations, deployment volume, customisation, security, hiring capacity, urgency, infrastructure, current stage and ongoing support.
How it decides
Five engagement models are scored against your answers with transparent rules. Close calls become a hybrid recommendation rather than a false certainty.
What you get
A recommended engagement model, the reasons, risks to manage, estimated team size and duration, required capabilities and providers that match them.
What you get
A result you can act on, with every number explained.
- Recommended engagement model
One of six models, with a secondary option where relevant.
- Estimated team size and duration
Indicative ranges derived from complexity and maturity, not a quote.
- Required capabilities
The capabilities a team must list to be a credible fit, linked to the directory.
- Matching providers
Providers scored against the profile your answers create, with reasons and gaps.
- An explanation, not just a score
Why this model, the risks to manage and concrete next steps.
- Your result by email (optional)
A link you can share with colleagues, plus occasional research.
The six possible recommendations
Each has a plain definition so you know what you are comparing.
- 1
Build internally
Your own engineers own the deployment. External help is limited to architecture review and evaluation design.
- 2
Hire a Forward Deployed Engineer
Recruit a permanent forward deployed engineer because repeated customer deployments justify an in-house function.
- 3
Staff augmentation
Specialist AI engineers join your team for a period, directed and managed by your own technical leadership.
- 4
Embedded FDE pod
A provider's forward deployed engineers embed in your environment and own delivery to production, transferring practice as they go.
- 5
External AI implementation partner
An external partner owns a defined scope end to end and hands over, with optional support afterwards.
- 6
Hybrid model
Two models scored almost equally. Combining them, for example hiring while an embedded pod bridges the gap, is the pragmatic route.
Frequently asked questions
- How long does the assessment take?
- About four minutes. There are twelve multiple-choice questions plus an optional question about what you are deploying.
- Do I need to give my email address?
- No. The result is shown immediately. Leaving an email is optional and is used to send you a link to the result and occasional deployment research.
- How is the recommendation calculated?
- Each answer scores twelve dimensions such as engineering capacity, AI expertise, integration complexity, urgency and ongoing support. The engine scores five engagement models against those dimensions and explains why the top one fits. If two models are within 1.5 points, it recommends a hybrid.
- Are the matching providers ranked by who pays?
- No. Providers are matched on stated capabilities, industry, technology, engagement model, region, team size, budget and availability. Featured placements are always labelled and never change the score.
- Is this a quote or a guarantee?
- No. The team size and duration are indicative ranges derived from your answers. Use the cost calculator and provider conversations to refine them.
Ready when you are
Twelve questions, about four minutes, no email required.