AI PoC preparation
![](https://www.aston.eu/wp-content/uploads/2023/03/agility-1024x778.webp)
We will help you to put together the necessary data and preconditions – whether or where the necessary data exist in your organization, how to annotate the data, how to frame the problem for AI PoC, definition of the good and wrong sample result.
Input:
- list of competent allocated people on the customer’s side
- definition of the problem to solve
Process:
- Workshops with outputs: use cases to solve are described in detail
- Workshops featuring: Lean Canvas
- Workshops to define hypotheses that we want to examine in the PoC
- Data preparation analysis: what data is available for PoC
- Definition of a good and bad results
- Data annotation – how to, how many
- Aston: data exploratory analysis
Outcome:
- presentation of results: use cases/lean
- canvas/hypotheses/data/exploratory analysis/problem framing
- our recommendation, whether it is possible to implement PoC and what expectations can be set
- our suggestion how to run the PoC – complexity and workload
AI PoC
1st model and tests
We will do a research and find the most suitable neural network model for a trial.