The problem this service answers
A project has been proposed internally, a budget is being prepared and a supplier is already involved in the discussion. Before committing, your management wants an assessment from someone with nothing to sell.
Projects on this subject rarely fail for the reason anticipated. The model is usually not the obstacle. What stops a project is data that turns out to be incomplete or unusable, a process that cannot absorb an automated output, a regulatory requirement discovered late, a cost structure that only appears at production volume, or an operating model nobody has agreed to run afterwards.
A feasibility study is worth most before the commitment is irreversible. Once budget is voted, a supplier is chosen and the project is announced internally, an unfavourable finding becomes politically expensive and the project continues regardless. Conducted early, the same study often reshapes the project into something smaller that succeeds rather than cancelling it.
What the assignment covers
We assess the project as described, against five dimensions: technical viability given the current state of the technology, the data conditions the project actually requires, the regulatory and contractual constraints that apply, the cost structure at production volume rather than at pilot scale, and the organisational conditions for the result to be used.
The work is documentary and analytical rather than experimental. We examine your project documentation, the technical claims made about the approach, comparable deployments where they can be documented, and the constraints specific to your sector. Where a decisive assumption cannot be verified from documentation, we say so and propose the targeted test that would settle it.
The conclusion is explicit. A project may be feasible as described, feasible under conditions that have to be met first, feasible at a different scale, or not feasible with the current state of the technology. We also state what would change the answer, since on this subject a negative assessment is often valid for a defined period rather than permanently.
How we work on it
- We examine the project as described, with the team that proposed it.
- We identify the assumptions on which the project depends, including implicit ones.
- We assess technical viability against documented capabilities rather than announcements.
- We review the data conditions, the regulatory constraints and the contractual dependencies.
- We examine the cost structure at production volume and the reversibility of the choices.
- We deliver an explicit conclusion, with the conditions and the elements that would change it.
Where the information comes from
Sources are selected with you at the start of the assignment and reviewed as the subject evolves.
- Your project documentation, technical specifications and available data descriptions
- Technical documentation, model cards and published capability evidence
- Regulatory texts and sector requirements applicable to the intended use
- Documented comparable deployments and published implementation feedback
- Independent evaluations, licence texts and provider service commitments
Possible deliverables
The format is chosen with you. A single assignment can combine several of them.
- A feasibility report with an explicit and argued conclusion
- An inventory of the assumptions the project depends on, each assessed
- A note on regulatory, contractual and data constraints
- A list of the conditions to be met, or the tests that would settle open questions
- A presentation of the findings to the committee deciding on the project
Who this service is designed for
What this service does not promise
A feasibility study assesses conditions and risks; it does not guarantee that a project judged feasible will succeed. Execution, data quality in practice and internal adoption remain determining factors that no prior study can settle.
How to start
Send us the project as it currently stands, including its objectives and its constraints. We propose an assessment perimeter and a delivery date before starting.