Machine learning has the potential to solve real-world problems. Ironically, the primary barrier to solving those problems is “data”. Limited, sparse, or high dimensional data can prevent ML models from producing meaningful results, and it is common for data scientists to spend the majority of their time addressing these data issues.
Here’s a startup that is striving to address these bottlenecks through a new machine learning algorithm – Machine learning startup Ensemble has developed “Dark Matter” which the company calls as ‘a new step in the data science pipeline’.
Dark Matter, according to the company, lowers the barrier to entry for data scientists to achieve state-of-the-art model performance. It is said to have the capabilities to produce superior results with limited or sparse data, simpler models, and less compute, opening the door to new modeling capabilities and applications.
Some of its use cases include forecasting including price predictions, supply & demand and Customer churn. It can also be leveraged for more accurate recommendations like ad placement, content suggestions and product personalization. Dark Matter promises to significantly optimize training by reducing the compute required as it can train on limited/sparse data.
The company recently raised $3.3M in seed funding, led by Salesforce Ventures with participation from M13, Motivate, and Amplo.
“This year, we launched a new embedding API that learns to approximate hidden relationships in data,” said Alex Reneau, CEO of Ensemble. “This foundational technology frees up data scientists to focus on experimentation and also makes ML viable for problems previously unable to be modeled, unlocking new capabilities for our customers.”
With this new funding, Ensemble is poised to accelerate its product development, expand its team, and reach customers across industries. The company is committed to pushing the boundaries of what ML can do and continuing to deliver frontier research in the field that translates into real-world impact.
