A Data Residency-as-a-Service platform founded in San Francisco in 2019 by Peter Yared, with roughly $50M raised (Series B, September 2025) and a team of about 59. It lets global SaaS applications such as Salesforce, ServiceNow, Okta, and Auth0 store and process sensitive data inside a specific country while the global app only sees cloaked or tokenised records, addressing data-localisation rules like China’s PIPL, Saudi SAMA, and EU GDPR. The mechanism is a digital twin per customer inside the target country, with a web services proxy and REST API handling create, read, update, delete, and search while the proxy handles cloaking and uncloaking. AgentCloak, launched September 2025, extends this to AI agents: it uses Qwen inside China to detect PII/PHI/PCI data and generalise it—for example, turning “headache” into “mild neurological condition”—and integrates via MCP Server or MCP Proxy, including real-time voice agents. Deployment options include multi-tenant on Alibaba Cloud, single-tenant on any in-country cloud, on-premise, and AWS Outposts. China operations run with Digital China and Alibaba Cloud, including support for MLPS 2.0 and CAC reviews. G2 shows two reviews averaging 3.5/5, with one reviewer praising API and SDK clarity while expressing uncertainty about PIPL compliance.
**Pros & Cons**:
Pros: Solves a concrete regulatory problem—keeping sensitive data inside a country while letting a global SaaS application operate—rather than a generic encryption or access-control story; the digital twin plus cloaking architecture means the global app never sees raw PII, which is the design regulators actually ask for; AgentCloak extends data minimisation to AI agents and supports MCP Server/MCP Proxy plus real-time voice, which is timely given agent adoption; China operations run with Digital China and Alibaba Cloud, including MLPS 2.0 and CAC support, which is not something a generic residency vendor can replicate; official Okta/Auth0 integration and AWS Outposts availability broaden deployment paths; roughly $50M raised with named investors gives it more operational stability than most early-stage infrastructure startups.
Cons: G2 review volume is extremely small—two reviews averaging 3.5/5, and one reviewer explicitly expressed uncertainty about PIPL compliance, so user validation is thin; pricing is completely opaque—there is no public pricing page, and AWS Marketplace requires contacting the seller; using it in China requires an ICP licence or filing, which is a regulatory prerequisite rather than a product limitation, but a real gate for teams entering that market; the AI angle is recent—AgentCloak launched in September 2025, so it should be described as a 2025 expansion rather than native AI infrastructure; the product’s centre of gravity is data-residency compliance, so it may not fit AI developers who do not operate in regulated markets; third-party competitor lists (Tracxn) misleadingly group it with storage infrastructure vendors like Nutanix and DataCore, which obscures its actual category.
Who is it for?:Global SaaS vendors that need to serve customers in China, Saudi Arabia, or the EU while meeting local data-localisation rules; enterprise architects and compliance teams implementing PIPL, SAMA, or GDPR data-residency requirements; organisations deploying AI agents that must minimise PII before sending data to models, especially inside China; teams already standardised on Okta/Auth0 that want residency controls without rebuilding their identity stack.