Ursa Computing offers a platform that optimizes data processing and collaboration for data science, machine learning, and AI teams, enabling faster model development and deployment. The solution targets inefficiencies in data handling, significantly reducing time-to-insight and improving team productivity.
Funding
$4.9M raised to dateRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
Founders
Product
Problem
Data science, machine learning, and AI teams face bottlenecks in data processing and orchestration, which slows down model development and deployment. Existing SQL engines often lack the performance and features required for modern AI workloads, particularly those involving GPUs and large datasets.
Solution
Theseus, by Voltron Data, is a GPU-accelerated and distributed SQL engine designed to optimize AI pipelines. It enables users to process data and drive Large Language Model (LLM) workflows on the same GPU, facilitating tasks from embedding and vector search to Retrieval-Augmented Generation (RAG) in DataFrames and SQL. Theseus supports in-place queries for data lakes, lakehouses, and warehouses, and is compatible with Apache Iceberg and standard file formats. Its Kubernetes-native control plane allows deployment across public and private clouds, ensuring security and compliance.
Target Audience
The primary target audience includes data scientists, machine learning engineers, and AI developers who require a high-performance SQL engine for AI workloads.
Features
- GPU-accelerated query execution for complex queries
- Support for User-Defined Functions (UDFs) to integrate custom logic into SQL
- Compatibility with Apache Iceberg and standard file formats
- Kubernetes-native control plane for deployment across various cloud environments
- Query Profiler for real-time analysis of query behavior at CPU, GPU, and I/O layers
- Integration with Python via Ibis API
- Conda install for local setup and cloud-based test drive environment