Sarus

Name: Sarus
Brand: Sarus

Privacy layer for Analytics & AI

Est 2019|Valuation $8.9M|Raised $2.7M

LinkedIn 2K|Website|FTEs 16|HQ Paris, Île-de-France, FR

MarketsData Governance Data Virtualization Data Security Platforms

01Competitive Landscape

Narrative analysis

#### Where Sarus appears differentiated - Use of DP combined with query rewriting as an enforcement layer across SQL and Python workflows is a distinctive architectural approach. - Documentation emphasizes integration with major warehouses and BI tools and a Python SDK for data science workflows. - The combination of DP, synthetic data, and query rewriting is positioned as a unified layer in vendor materials. #### Where alternatives may be preferred - Decentriq: may be better aligned for partner ecosystem clean rooms and audience activation. - Duality: may be preferable when a broader PET portfolio (homomorphic encryption, federated learning) is required for collaboration with specific cryptographic guarantees. - Platform-native governance (Snowflake/Databricks): if the environment is tied to a single data platform and needs access controls, cataloging, masking, and policy enforcement within that platform, native tools may be operationally simpler though they may not provide DP guarantees. #### Market adoption signals - Participation in startup accelerators and selection for grant/accelerator programs indicate R&D investment and external validation. - The absence of visible review corpora on mainstream platforms suggests early-stage adoption, limited public reviews due to regulated use cases, or limited presence on review sites.

	Sarus (vendor site)	Decentriq (vendor site)	Duality (vendor site)
Starting Price	Published per-dataset pricing for tiered datasets; platform license fee available by quote	Quote-based (not publicly listed)	Quote-based (not publicly listed)
Best Fit	Organizations seeking DP-enforced analytics/ML and computation-to-data workflows, with synthetic data capabilities (vendor positioning)	Cross-company data collaboration and clean-room use cases; audience/marketing ecosystem emphasis	Privacy-enhancing technologies (homomorphic encryption, federated learning, TEEs) for secure collaboration across organizations
Key Strength	DP-centric privacy enforcement layer; explicit dataset pricing for published tiers; support for SQL + Python + BI connectors + synthetic data narrative	Clean-room network and audience activation features; some competitors publish SOC2 Type 2 certification	Broad PET toolkit and enterprise narratives for cross-border collaboration; multiple PET modalities available
Key Weakness	Limited public third-party reviews; enterprise security attestations not clearly public; DP tuning complexity and accuracy trade-offs	Marketing and audience-centric positioning may be less focused on internal analytics privacy layers	PET approaches can be complex, costly, and incur performance/latency trade-offs depending on the technique

01Competitive Landscape

Narrative analysis

Where Sarus appears differentiated

Use of DP combined with query rewriting as an enforcement layer across SQL and Python workflows is a distinctive architectural approach.
Documentation emphasizes integration with major warehouses and BI tools and a Python SDK for data science workflows.
The combination of DP, synthetic data, and query rewriting is positioned as a unified layer in vendor materials.

Where alternatives may be preferred

Decentriq: may be better aligned for partner ecosystem clean rooms and audience activation.
Duality: may be preferable when a broader PET portfolio (homomorphic encryption, federated learning) is required for collaboration with specific cryptographic guarantees.
Platform-native governance (Snowflake/Databricks): if the environment is tied to a single data platform and needs access controls, cataloging, masking, and policy enforcement within that platform, native tools may be operationally simpler though they may not provide DP guarantees.

Market adoption signals

Participation in startup accelerators and selection for grant/accelerator programs indicate R&D investment and external validation.
The absence of visible review corpora on mainstream platforms suggests early-stage adoption, limited public reviews due to regulated use cases, or limited presence on review sites.

	Sarus (vendor site)	Decentriq (vendor site)	Duality (vendor site)
Starting Price	Published per-dataset pricing for tiered datasets; platform license fee available by quote	Quote-based (not publicly listed)	Quote-based (not publicly listed)
Best Fit	Organizations seeking DP-enforced analytics/ML and computation-to-data workflows, with synthetic data capabilities (vendor positioning)	Cross-company data collaboration and clean-room use cases; audience/marketing ecosystem emphasis	Privacy-enhancing technologies (homomorphic encryption, federated learning, TEEs) for secure collaboration across organizations
Key Strength	DP-centric privacy enforcement layer; explicit dataset pricing for published tiers; support for SQL + Python + BI connectors + synthetic data narrative	Clean-room network and audience activation features; some competitors publish SOC2 Type 2 certification	Broad PET toolkit and enterprise narratives for cross-border collaboration; multiple PET modalities available
Key Weakness	Limited public third-party reviews; enterprise security attestations not clearly public; DP tuning complexity and accuracy trade-offs	Marketing and audience-centric positioning may be less focused on internal analytics privacy layers	PET approaches can be complex, costly, and incur performance/latency trade-offs depending on the technique