Amplifier AI onboarding process starts with an integration meeting to review data infrastructure, agree on data fields, and share a data sample. After validating and transferring data, Amplifier AI models are trained and deployed for live use.
Client Review and Scheduling
As a preparation for onboarding , we advise reviewing the Data Integration guide to respond with any questions or clarifications needed.
Step 1: Integration Meeting
Meeting topics:
- We will review together your data infrastructure and schemas – it is recommended to provide your schema documentation.
- We will agree on the data fields needed for the transmission.
- We will agree on the transmission mechanism – selected integration method, security, frequency (real-time/batch) - depending on your abilities and the products purchased.
- We will agree on the historical data to be shared by the client (6 months preferable, 3 months minimum). This is needed for Amplifier AI’s models training set.
Meeting output: Access method is shared with the client, this will enable you to share a data sample, using the agreed transmission mechanism.
Step 2: Data Sampling
- Client shares a data sample (typically covering a few days of activity), using the agreed mechanism.
- Anthem reviews and confirms together with you the data is indeed complete.
Step 3: Data Transfer
- Daily/real-time connection is created and tested.
- Daily data is transmitted from the client.
- Anthem reviews data and compiles an investigation report for your review.
Step 4: Data transformation
- Anthem validates the rules and mapping logic applied correctly.
- The client transfers the historical data.
- Anthem reviews data and compiles an investigation report
Step 5: Live Release
- Amplifier AI models are trained on client’s data.
- Amplifier AI models outputs undergo iterative evaluation and fine-tuning to enhance accuracy.
- Amplifier AI is released to client, relevant modules are activated!