Deep AI for
Speed up scientific discovery with AI image analysis.
Join world-class researchers and scientists across the globe.
Train AI from your images, run in one click. No code required.
Biodock makes deep AI easy to train, run, and interpret, radically improving the accuracy and consistency of image analysis.
Create a model in 15 minutes of work
Biodock's AI-assisted labeling allows you to build and improve powerful analyses in a fraction of the time.
AI-assisted labeling and improvement
Use your previous models to automate 95% of labels to drive iterative improvement.
Activate and customize your metrics
Once finished, you're ready to run on new images. Customize metrics around area, intensity, colocalization, and more.
Your model automates analysis in one click
Your AI pipelines run in a single click from our distributed platform built for scientists, delivering the comparisons you need at any scale.
Blazing fast at any scale
We run your images in parallel on large clusters to achieve up to 3000x acceleration on larger jobs. Start multiple jobs at a time.
Built-in comparison of experimental groups
Compare experimental differences with beautiful custom plots and object-level metrics. Filter subgroups like flow cytometry.
Interactive object viewer and editor
Our interactive viewer allows you to visualize detected objects with overlays and dive into quantitative metrics like area and signal intensity.
Easy to use and understand
No configuration, installation, or code. Biodock's clear online docs and intuitive interface makes onboarding easy for any scientist.
See the platform in a quick demo
See how Biodock works in a walkthrough demo, covering training a model, running it, and configuring it for different scenarios.
End-to-end deep AI platform for scientists
Connect data from anywhere, and easily label and train to create powerful quantitative AI modules. Our powerful compute execution and results dashboard help you run and interpret your analysis quickly.
Organize your files in one place and connect images in from cloud providers like S3, or Drive, OneDrive, Box, or Dropbox.
Biodock allows for seamless zoomed viewing on multi-GB images and file sharing with one click.
Collaboratively label your images in our AI Model Dashboard to quickly put together a training set.
Then, convert it into a versioned AI pipeline custom built for your images by clicking Train.
Biodock runs across massive clusters to rapidly and reliably run any size of job up to a thousand times faster.
Just click run on your trained AI module, with zero parameters, code, or configuration.
Results drop into our Dashboard, where they are ready for data and graph export, as well as powerful population grouping and statistical comparison.
For every object detected - get quantitative metrics like area, x-y position, orientation, roundness, and intensity expression.
Fit for any biological images
Train AI with any type of 2D image: human, animal, prokaryote or inorganic. Supports multichannel images with up to 100 channels. Flexible enough to adapt to any target image.
Single mitochondrial identification from transmission electron microscopy images.
Single human cell identification from cultured dish brightfield images.
Nerve fibers and epidermal layer identification in human IF skin biopsy images.
Intestinal villi and erosion identification in murine H&E stained intestinal tissue sections.
Collagen stained region identification in human breast cancer tissue.
Brown-stained infiltrated immune cell identification in human breast cancer tissue.
Biodock labeled and trained a custom AI model for us, achieving human-level performance and allowing us to expand our image analysis pipeline.
Biodock enabled us to provide controlled access to our analysis for our scientific partners, enabling them to easily analyze their results on the cloud without distributing our internal algorithms.