Provides a privacy-preserving analytics and AI platform that enables data scientists and analysts to query sensitive data without direct access, using differential privacy, synthetic data generation, and privacy-first query rewriting. This approach ensures compliance with data protection regulations, prevents breaches, and allows organizations to unlock the full value of their data while maintaining security and privacy.
Funding
$2.7M 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
Organizations struggle to unlock the full potential of sensitive data for analytics and AI due to privacy regulations, security concerns, and the risk of data breaches. Traditional anonymization techniques like data masking and hashing are often insufficient to prevent re-identification, leading to compliance issues and limiting data utility. Sharing data across silos and with external partners further exacerbates these risks.
Solution
Sarus provides a privacy layer that enables data scientists and analysts to query sensitive data without direct access, ensuring compliance and maximizing data value. The platform leverages techniques like differential privacy, privacy-preserving synthetic data generation, and privacy-first query rewriting to protect data while preserving its analytical utility. Sarus integrates with existing data stacks and cloud environments, allowing organizations to perform research, analytics, and AI safely. By providing provable privacy guarantees, Sarus allows organizations to collaborate on sensitive data, accelerate project timelines, and unlock new insights without compromising data security or regulatory compliance.
Target Audience
Sarus targets data security teams, data protection teams, and AI & analytics teams within organizations in industries such as finance, healthcare, marketing, and the public sector.
Features
- Differential privacy implementation to ensure query results are provably safe
- Privacy-preserving synthetic data generation using generative AI models trained with differential privacy
- Privacy-first query rewriting to transform data processing jobs into privacy-safe versions on-the-fly
- Integration with Azure Confidential Clean Rooms for secure multi-party data collaboration
- SarusLLM, enabling the use and fine-tuning of Large Language Models (LLMs) with private data safely
- Qrlew, an open-source library designed to rewrite SQL queries into privacy-safe queries
- Support for fine-tuning LLMs with differential privacy in Databricks
- Output-level controls to ensure privacy compliance for all query outputs
- Integration with existing data tools via SQL, pandas, and sklearn