Valohai Brings AI to GitHub
Speed to market enables first-mover advantage and drives revenue.
Join the DZone community and get the full member experience.Join For Free
We had the opportunity to meet with Eero Laaksonen, CEO of as part of the IT Press Tour in San Francisco. Valoh is the Finnish word for lantern shark, a deep-dwelling, self-illuminating fish.
Eero believes automation is key to improving quality of life and that deep learning makes things that are not scalable today, scalable like looking at medical images and autonomous cars. Valohai strives to push all industries forward with the ability to do meaningful things faster.
You might also like: How to Do Deep Learning for Java
There is an issue with the usability of machine learning (ML). No longer is it an issue of models, it is more about making it easy to put the models into production. Deep data sets on huge models make it trivial to work on these. Like software and application development, there is now demand for higher quality, better-performing models to be developed faster.
Valohai is providing tools to do ML faster because speed drives revenue. The ability to detect cancer cells from a medical image so it can scale and generate billions. Speed to market gives a great competitive advantage. Self-driving cars are a race to who gets there first. There is a race to see how fast you can put scalable models in the market.
Valohai has a recipe for production scale deep learning and ML that address the inherent problems of experiment reproducibility and regulatory compliance. Today, you need to be able to explain how decisions are made and how the model works. GDPR requires traceability through the process to be able to remove customer data when the request is made.
Valohai enables the fast onboarding of teams for companies that need to grow fast to get value by knowing what models are running and who’s working on what. Data scientists frequently leave a job and the company needs to know what they’ve done and that they have been following testing and documentation procedure. Valohai provides a full audit trail of every experiment that has been done.
That’s important because deep learning and ML requires many quick experiments:
- Trial and error to know how much data is needed to get to the answer
- Put more hardware to get results from more data faster
- Have to run through a TB of data to know if a change worked
Valohai provides an ML platform as a service with everything being open in terms of platforms and tools. There is an enterprise focus, and customers own data
Machine vision is a core use case for the solution. A Canadian company is monitoring toxic online content and collaborating with Canadian Homeland Security to stop child pornography.
Drones are detecting the maintenance needs of electric lines and transformers.
Other use cases include automation of preventive and predictive maintenance in the power grid. Prediction models in finance. Future load forecasts for Telco’s to build new towers to meet demand and real-time optimization of networking paths for time of day.
Applying NLP in Java: All From the Command-Line
Opinions expressed by DZone contributors are their own.
A Complete Guide to AWS File Handling and How It Is Revolutionizing Cloud Storage
RBAC With API Gateway and Open Policy Agent (OPA)
Observability Architecture: Financial Payments Introduction
How to LINQ Between Java and SQL With JPAStreamer