Let’s Train Something Great

Data science to power any product

When your product needs a high performing algorithm, powerful computer vision, or intelligent machine-learning AI, we can help develop or improve your offering.

Some Success Stories

Outcome

Bloom is revolutionizing health tech with daily urinalysis to track your 15+ important medical bio-markers: Ketones, Vitamin-C, Hydration, etc.

The Bloom computer vision algorithm wasn’t performing where the team needed to reach their next inflection point, raise follow-on funding, and launch to the masses. Our team worked with them to dramatically improve their technology, add additional functionality, and get to the next level.

Improvement

+190%

improvement to existing algorithm

Outcome

5k+ pre-orders of 40k+ products. Raised multiple rounds of venture capital.

Outcome

Best Apps is a custom apparel platform, where fans can partner with their biggest stars - whether celebrities, brands or athletes - to create custom clothing.

Best Apps needed our help developing a moderation system to monitor and approve images and custom text being added by fans. Uploading your own image next to famous Mohammed Ali was a huge innovation, and they needed our help to execute it. Our team developed their entire machine learning moderation strategy and built it scalably.

Performance

96%

moderation accuracy (4% elevated to human review)

Outcome

Major launch with A-list celebrity. 4m visits in 3 days over $1m in revenue. Raised large Series A.

Outcome

Streamoid is a fashion-tech company who’s super-intelligent fashion AI acts as a highly trained stylists and shop-assistants while providing the largest fashion retailers with an in-depth knowledge of their product catalogue.

Streamoid needed help training their machine learning models, including building the models from scratch and sourcing training data. Our team built the scraping engine to source and label training data and then trained a recursive neural network to power Streamoid’s AI.

Performance

95%

garment classifier across 100+ categories

Outcome

Partnership with Target Raised multiple rounds of venture capital

Our data scientists - from places like Stanford, CalTech and MIT - are highly collaborative, efficient and pragmatic.

partner logopartner logopartner logopartner logo
Ed

Ed

HEAD OF DATA SCIENCE

Data scientist with 15 years experience using statistics and machine learning to analyze datasets in physics, finance, and online petitions. B. Magna Cum Laude in Chemistry & Physics from Harvard College and Ph.D from Cal Tech in Experimental High Energy Particle Physics.

Darin

Darin

SENIOR DATA SCIENTIST

Fluent in Agile principles, specializing in mathematical modeling, statistical analysis, and feature extraction from large data sets. PhD in Physics (quantum computing), Stanford. BA in Physics, Princeton. First to initialize and optically detect the spin of a single electron trapped within a F:ZnSe semiconductor system.

Jonatan

Jonatan

DATA SCIENTIST

Computer Vision specialist with high affinity for Deep Learning solutions since they broke out in 2012. Technical skills include TensorFlow/Keras, OpenCV, Houdini FX and other Machine Learning / Computer Vision / CGI frameworks in both C++ and Python.