Products
Get Access
Where AI meets
the real world.
Bring ambitious, sensor-driven machine-learning products into production through integrated data-collection and testing.
Get Access
Our Customers and Partners
Production is where ML goes to die. Successful real-world ML requires a closed-loop.
Today, AI too often is built in isolation. Teams build models, deploy them to the edge but have little insight into how they perform in the real world and if they work for their users in every edge-case.
An E2E platform for Data Collection, Model Development, and Testing.
Dashboard
All your data, in one place.
Manage all your data in one place from day one. From manually collected data to field-data, FieldDay means no more tedious looking through S3-buckets.
Collaborative timeline interface.
FieldDay lets you look at collected events on a timeline alongside annotations and analytics from the field making it simple to understand what is going on.
Model inference.
FieldDay lets you load any model and run it over field data, enabling powerful analysis and automatic detection of false-positives.
Hardware Integration
Run on your hardware.
Integrate into your existing hardware infrastructure. Start emitting FieldDay data events and you will start to see data show up in the dashboard.
Manage your device settings.
Manage device settings from the FieldDay dashboard, giving your modelling team control over the quality and consistency of the data being collected.
FD StarterKit
If your hardware is not ready yet we have assembled production-ready hardware in a neat little kit that you can start collecting data with today.
Ready-to-go Collection Tools
Real-time, interactive labelling
Annotating time-series data after the fact is hard. Our real-time annotation interface allows your collection team to annotate data while they collect.
Connects with your hardware
The FieldDay SDK on your hardware will automatically be available through our Collect Interface on Android or iOS.
Configurable + live metadata
Add custom metadata before starting a collection session. API-driven data like weather or location is updated automatically and continuously.
Enabling a 105M$ joint-venture by Ford and ADT
to bring cutting-edge ML
to production.
Case Study Coming Soon
Frequently asked questions
Is FieldDay a training pipeline?
No, but it probably integrates with yours! FieldDay focusses on the training data itself, not how your models are trained. Therefore we designed it to seamlessly integrate with your existing infrastructure.
Does FieldDay work with GCP and AWS?
From the ground up we designed FieldDay to be agnostic about cloud infrastructure. You can bring whatever setup you are running with and should be able to set up FieldDay within those constraints.
Aren’t there already so many tools?
Yes, there are many tools, especially when it comes to data labelling and model training. This is why FieldDay focusses on data collection at the edge, a niche we find painfully underserved.
Who should use FieldDay?
We believe any team that wants to develop an intelligent sensor should use FieldDay. Be it for IoT at home, an assistive driving system or even data that is collected on the server. If it’s time-series, FieldDay can help.
Will this work with my hardware?
We have built the FieldDay hardware SDK to be as simple as possible. While you will need to write a few lines of code to integrate it a lot of the heavy lifting of data collection is done for you.
Just image, audio and motion?
While we find that those sensors cover the use-cases of most of our customers, you can extend the FieldDay timeline with your own visualisations while still profiting off of everything else the platform has to offer.