Vertex AI · BigQuery · Cloud Storage · Cloud Run · GKE · IAM
Training, storage, and serving sit in one place.
Google Cloud
The models live on Google Cloud.
Training, history, files, and the API are one stack. This page is the architecture we are building. It is not a claim that every piece is already in production.
Company active since 27 June 2025.
Why this stack
The product is the model plus the path around it.
Google Cloud
Vertex AI
Train, evaluate, pin a version, serve.
BigQuery
The record.
Cloud Storage
Raw intake.
Cloud Run
The endpoint the dashboard calls.
GKE
For a longer run or heavier serving.
IAM
Identity on every door.
How a reading moves
Path of one reading
From the file to the line on the screen.
- Cloud StorageThe file hits Cloud Storage with the plot id.
- BigQueryThe pipeline writes a row in BigQuery and keeps what it dropped.
- Vertex AIVertex AI serves the pinned version for that crop.
- Cloud RunCloud Run returns the slip, or a miss.
- DashboardThe dashboard shows the line in the operation’s words.
- AdjustThe walk note returns and waits for the next version.
The training loop
A new version starts from a decision sentence, a data cut, and the misses of the version before it.
Built, not badged
Fitocrops is a Google Cloud architecture.
What the grower does not open
Jobs
Training runs live behind the version.
Versions
The old one stays addressable.
Doors
IAM on storage, tables, and the endpoint.
The miss log
Every reading that did not fit stays.
How the money is shaped
The farm pays a subscription for the software.
What we will ask before a version
- The crop and the decision sentence.
- Which files already exist.
- Who is allowed to see which farm.
- What a miss should say.
Contact
Bring the crop and the data you already have.
We answer with the fit, the gap, and whether a first version makes sense. Terms come after that, not before.