Security Review FAQ
This page is designed for IT security teams, procurement officers, and risk reviewers evaluating Sensei IQ as part of a vendor security review. The questions below reflect the most common inquiries we receive.
If you need a structured document for your review process, you can download our Pre-Filled Vendor Security Questionnaire — a complete Q&A document formatted for vendor security review workflows.
Before reviewing: Sensei IQ is not a traditional SaaS product. It is deployed as a Power Platform solution inside your own Microsoft 365 tenant. Many standard vendor security questions assume a vendor-hosted environment — in Sensei IQ's case, the "environment" is your own. We've structured these answers to reflect that reality clearly.
Connected Intelligence is in internal beta and is not yet activated for any client environment. Client activation is expected late 2026. The description below is the data-handling design as it will ship.
Company & Product
What is Sensei IQ and how is it deployed?
Sensei IQ is a project and portfolio management solution built on Microsoft Power Platform. It is deployed as a Power Platform solution into your organization's own Dataverse environment. There is no Sensei-hosted service receiving or storing your data — the entire application runs inside your Microsoft 365 tenant.
Who is Sensei Project Solutions?
Sensei Project Solutions is a Microsoft Solutions Partner specializing in project and portfolio management on Microsoft Power Platform. We develop and maintain Sensei IQ and are responsible for the security of the code and configuration we deploy.
Data Storage & Residency
Where is my data stored?
All Sensei IQ data is stored in Microsoft Dataverse within your own Power Platform environment. Your data resides in the geographic region you selected when provisioning your environment. Sensei does not receive copies of your data, and it does not leave your tenant.
The one exception is the optional Connected Intelligence AI service: when it becomes available and an administrator enables it, running an advisor sends a defined, scoped set of fields out for AI processing. All Sensei-managed infrastructure for this service runs in US-region Azure (West US). This applies only when the service is enabled and invoked — see the Connected Intelligence Advisors FAQ.
Is the data sent to Connected Intelligence anonymized?
No. Data sent to an advisor is not anonymized, masked, or pseudonymized. The control is minimization: only an explicit, named allow-list of fields leaves the tenant. If a listed field contains a personal name, it is sent as-is. The Connected Intelligence Advisors FAQ sets out what each advisor sends and where its field list is held.
Our contract requires data to stay inside our own tenant. Can we use Connected Intelligence?
If your contract requires data to remain within your own tenant, the Sensei-managed option does not meet that requirement. The self-hosted (Bring Your Own Model) option does, and is targeted for early 2027. Everything outside Connected Intelligence already runs entirely inside your tenant.
Does Sensei have access to my data?
Under normal operating conditions, Sensei Project Solutions has no access to your Dataverse data. Sensei staff access is only possible if your tenant administrators explicitly grant it for a specific purpose (such as implementation support), and it can be revoked at any time.
What is your data retention policy?
Data retention for Sensei IQ data held in Dataverse is governed by your organization's Microsoft 365 and Dataverse retention policies — not by Sensei. You configure and own data retention within your tenant.
The optional Connected Intelligence service is the one path where data is handled outside your tenant, and it carries its own retention terms. Sensei retains no copy of the scoped request. Anthropic deletes API inputs and outputs within 30 days by default. No Zero Data Retention agreement is in place for this service.
Encryption
How is data encrypted at rest?
All data in Dataverse is encrypted at rest using AES-256 encryption managed by Microsoft. Optionally, your organization can configure customer-managed encryption keys (CMK) through the Power Platform Admin Center.
How is data encrypted in transit?
All data in transit is encrypted using TLS 1.2 or higher, enforced by Microsoft Power Platform. This applies to all communication between users, the Power App, Dataverse, and integrated Microsoft services.
Who manages encryption keys?
By default, encryption keys are managed by Microsoft. Your organization can optionally take ownership of keys through the CMK feature in Power Platform.
Identity & Access
How do users authenticate to Sensei IQ?
Authentication is handled entirely by Microsoft Entra ID (formerly Azure Active Directory) using OAuth 2.0. Sensei IQ has no separate identity store. All of your organization's existing identity policies — MFA, Conditional Access, sign-in restrictions — apply automatically.
Does Sensei IQ support multi-factor authentication (MFA)?
Yes. MFA is enforced through your organization's Entra ID Conditional Access policies. Sensei IQ does not bypass or override your MFA configuration.
How is access controlled within the application?
Authorization is managed through Dataverse security roles, which your administrators assign and manage. Sensei IQ ships with pre-defined roles designed around least-privilege principles. Role management is entirely within your organization's control.
How is access controlled for Connected Intelligence?
In the advisors available today there is no prompt entry surface for end users. A user cannot alter the model's instructions or widen the field scope — both are administrator-held settings. The model has no write access to Dataverse. A suggestion persists only when the user accepts and saves it, which records an ordinary field update by that user.
Does Sensei use service principals in our tenant?
Sensei IQ uses two applications — Sensei IQ Deployment (installs and updates the managed solution) and Sensei IQ Applications (powers the client experiences, including the Teams app). Neither holds any application permissions; every permission on both is delegated. Both are:
- Registered in your tenant (visible and manageable in your Entra ID portal)
- Granted only the minimum permissions required
- Audited through your tenant's logs
- Revocable at any time by your administrators
Both Application IDs, every Microsoft Graph scope they request, and why each one is needed are published at Permissions & External Endpoints so your administrator can compare the list directly against the consent screen.
What external endpoints does Sensei IQ touch, and can they be turned off?
Every endpoint outside your tenant is listed at Permissions & External Endpoints, with what data each carries and how to disable it. In summary: the deployment pipeline and our static asset host carry no customer data at all; diagnostic telemetry and in-app feedback are optional and each disabled with a single setting; and Connected Intelligence is opt-in, off by default, and not yet activated for any client environment. With those three off, no customer data reaches any Sensei-operated system.
Network Security
What network ports or inbound connections does Sensei IQ require?
Sensei IQ does not require any inbound network connections to Sensei-operated infrastructure. It operates entirely within Microsoft's Power Platform trust boundary using standard M365 HTTPS/TLS connectivity.
Does Sensei IQ expose any APIs or webhooks on the internet?
No. Sensei IQ does not operate internet-exposed endpoints. All communication is outbound from within your tenant to Microsoft services.
Application Security
What security standards does Sensei follow in development?
Sensei follows the Microsoft Security Development Lifecycle (SDL). All code changes require peer-reviewed pull requests. Automated scanning (CredScan, BinSkim, ESLint, CodeQL) runs on every build. See Development Practices for details.
Has Sensei IQ undergone third-party security testing?
Sensei IQ undergoes regular static analysis scanning. Clients who wish to conduct security testing against Sensei IQ in their own environment are welcome to do so — because Sensei IQ runs in your tenant, you can engage a third-party assessor to test your own environment without requiring Sensei's involvement or approval.
Is Connected Intelligence exposed to prompt injection?
In the advisors available today there is no prompt entry surface: a user invokes a fixed, curated instruction over a fixed field list and cannot alter either. The model has no tool access, no network or database access, cannot execute actions, and cannot write to Dataverse. Output is plain text placed into one named field pending human acceptance. Advisors that accept user-supplied content are assessed separately before activation.
Is Sensei IQ affected by the PHP vulnerabilities disclosed in October 2026 (CVE-2026-109618 / CVE-2026-109619)?
No. These vulnerabilities, sometimes called "php2shell", affect only systems that run the PHP interpreter. Sensei IQ does not use PHP in any part of the product:
- Platform: Sensei IQ runs inside your Microsoft 365 tenant on Microsoft Dataverse, Power Apps, Power Automate, and Power BI. Microsoft operates and maintains all of these services.
- Product code: Sensei IQ's user interface controls are written in TypeScript/JavaScript, and its Dataverse business logic is written in C#/.NET.
- Connected Intelligence: when the optional Connected Intelligence service is available and enabled, its Sensei-managed Azure Function runs on Node.js.
No action is required from your organization for Sensei IQ. If you are tracking vendor responses, you can record Sensei IQ as not impacted.
Does Sensei use AI in the product or in development?
Sensei IQ offers an optional, opt-in AI capability called Connected Intelligence, which is off by default and is not yet activated for any client environment. When it becomes available and an administrator enables an advisor and a user invokes it, a defined, scoped set of fields is sent over an encrypted connection to a Sensei-managed Azure Function and on to Anthropic's Claude API, which returns a suggestion the user can accept, edit, or reject. Under Anthropic's commercial terms, "Anthropic may not train models on Customer Content from Services," and commercial API inputs and outputs are not used for training by default; you retain rights to your inputs and own the outputs. Anthropic deletes API inputs and outputs within 30 days by default, and no Zero Data Retention agreement is in place for this service. When Connected Intelligence is not enabled, no product data leaves your tenant for AI processing. See the Connected Intelligence Advisors FAQ for the full data-handling detail.
Separately, Sensei's development team uses productivity tools including GitHub Copilot, Claude, and Microsoft 365 Copilot. All AI tool usage in development is governed by Sensei's AI Policy, which requires human review of AI-generated output before it is committed.
Compliance & Certifications
Does Sensei have a SOC 2 report?
Sensei Project Solutions does not independently hold a SOC 2 certification. However, because Sensei IQ runs on Microsoft Power Platform and Dataverse, it operates within infrastructure that is covered by Microsoft's SOC 2 Type II certification. Microsoft's compliance reports are available through the Microsoft Service Trust Portal.
What compliance frameworks does Sensei IQ support?
Through Microsoft's platform, Sensei IQ operates within an environment that supports SOC 1/2, ISO 27001, ISO 27018, GDPR, HIPAA/HITECH (with a Microsoft Business Associate Agreement), FedRAMP, and more. The applicable coverage depends on your organization's Microsoft 365 licensing and configuration. See Compliance for more detail.
Is Sensei IQ GDPR compliant?
Sensei IQ is designed with GDPR principles in mind. Your organization is the data controller for Sensei IQ data. For the base product Sensei is not a data processor — Sensei IQ operates within your environment and Sensei does not receive your data. GDPR controls (data minimization, retention, right to erasure) are exercised through your Microsoft tenant tools.
When the optional Connected Intelligence service becomes available and an administrator enables it, Sensei acts as a processor for the scoped fields sent on each advisor request, with Anthropic as a subprocessor; data minimization is preserved because each advisor sends only a defined, named set of fields. Those fields are minimized, not anonymized — see the Connected Intelligence Advisors FAQ.
Incident Response & Business Continuity
How does Sensei respond to security incidents?
Sensei maintains a documented incident response process covering detection, containment, escalation, assessment, remediation, notification, and post-incident review. For incidents involving client tenant data, the primary response capability lies with the client's own security tools (Microsoft Purview, Defender, Entra ID). See Incident Response for details.
For Connected Intelligence, the same process adds two AI-specific containment steps. Any advisor can be disabled immediately via its configuration setting, a kill switch that requires no deployment, and the model credential can be rotated independently. Because Sensei holds no persistent copy of client data, exposure is limited to in-flight requests. Disabling the Insights advisor also stops its scheduled preparation, not only its button.
What is Sensei's business continuity and disaster recovery plan?
Sensei maintains quarterly release backups including source code, deployment artifacts, CI/CD pipeline definitions, and release documentation. This enables recovery from build infrastructure failures, repository incidents, or the need to roll back to a known-good release state.
For client data, business continuity and disaster recovery is handled through Microsoft Power Platform's built-in environment backup and point-in-time restore capabilities, which your administrators control.
What is Sensei's uptime SLA?
Sensei IQ's availability is governed by Microsoft's Power Platform service SLAs. Sensei does not operate infrastructure with a separate uptime commitment.
Vendor Risk & Subprocessors
Does Sensei share my data with subprocessors or third parties?
By default, no. Sensei does not receive, store, or share your organizational data, and for the base product there are no subprocessors — Sensei IQ's only dependencies are Microsoft platform services (Power Platform, Dataverse, Azure) operating within your own tenant.
The one exception is the optional Connected Intelligence service. When it becomes available and an administrator enables it, running an advisor sends a scoped set of fields to be processed by a Sensei-managed Azure Function (with Sensei acting as a processor) and Anthropic's Claude API (acting as a subprocessor), for that request only. The subprocessors used exclusively for this optional service are Anthropic, PBC (AI model provider — United States) and Microsoft Azure (Sensei-managed hosting — Azure West US). This applies solely when the service is enabled and invoked. See the Connected Intelligence Advisors FAQ.
What cloud providers does Sensei use?
For Sensei's own internal operations (development, source control, email, corporate systems), Sensei uses Microsoft 365 and GitHub Enterprise. The base Sensei IQ product is delivered entirely on Microsoft platform services. When the optional Connected Intelligence service becomes available and an administrator enables it, Sensei additionally operates an Azure Function that calls Anthropic's Claude API; all Sensei-managed infrastructure for this service runs in US-region Azure (West US). This applies only when that service is enabled and invoked.
Could Sensei change AI providers or subprocessors?
The architecture is model-agnostic. Any change of model provider or subprocessor would be communicated in advance and would remain under the same data-handling commitments: no training on client data, and no persistence at Sensei.
Right to Audit & Penetration Testing
Can we audit Sensei IQ?
Because Sensei IQ runs in your own tenant, your organization's auditors have direct access to the relevant logs and controls via Microsoft Purview, Power Platform Admin Center, and Entra ID.
Can we conduct a penetration test?
Because Sensei IQ is installed in your tenant, your organization can engage a third-party assessor to test your own Power Platform environment. This does not require Sensei's involvement — you own the environment. Sensei does not conduct penetration tests against client environments, as each client's security configuration is unique and those tests are most meaningful when scoped to the client's actual environment.
Download
For a printable version formatted for vendor review workflows, download the Sensei IQ Vendor Security Questionnaire (DOCX).