Research & labs
Instrument data, reproducibility and funder obligations evolve during a project. I build with the team the data platform, the provenance, the access and the compute that turn results into an institutional asset.
Needs I take on
A chain from raw instrument file to published figure, versioned end to end, so any result regenerates on demand
The funder’s data-management plan delivered as running infrastructure: repositories, formats, access and retention live with the project
Models over the unpublished corpora on a local architecture sized for the work: machines, models, permissions and logs designed together
What I put in place
- Pipelines with an owner, versions, controls, provenance and recovery procedure, from the management plan to the selected repositoryData →
- An AI architecture assessed against confidentiality, contract, transfer, compute, logging and scientific-validation constraintsArtificial intelligence →
- A storage architecture sized by access, growth, integrity, recovery, retention, exit and full costAzure and Google Cloud →
To go deeper
Test the fit: Research & labs
Describe the context, constraints and decision you need to make. The first conversation qualifies scope, boundaries and the next useful step.
Describe the situation