Dovetail 发布 Sun’s Out 企业级治理、上下文工程与标准化能力

Our Sun’s Out launch brings Digital Twins, Agents, a more connected ecosystem, and now enterprise-grade governance and security capabilities. Here’s what it means for your organization.

Grounding AI in what your business knows

A general-purpose AI tool answers from whatever it was trained on. It doesn’t know your company’s positioning, the tone your brand guidelines require, or which customer segment a question is actually about. When a model has no context to work from, hallucinated answers and off-brand responses are a given.

Our context engineering controls fix that by giving you two layers to set it at:

• Global workspace context—set once and it’s applied across every Project and Channel in your Dovetail workspace. This is where company strategy, brand voice, and compliance standards live, so answers stay consistent no matter which team is asking.
• Local Project or Channel context—scoped to a single Dovetail Project or Channel for the detail that’s relevant there such as a specific customer segment, a product line, or a research question a team is actively chasing.

Setting context controls how the AI responds so answers stay on brand. It reduces bias and hallucination because the model is reasoning from context you defined, rather than guessing. And it gives every answer strategic business context, so a question about churn or a competitor comes back framed the way your business thinks about it, not the way a general model would.

Every answer Dovetail’s AI gives comes with rich, traceable citations back to the real customer call, ticket, or session it drew from, so everything is grounded in real evidence.

Deploying AI without losing control

Governance controls what happens to your data, who has access to it, and is the compliance layer that lives up to your security team’s scrutiny.

AI redaction protects sensitive data automatically. Admins define what counts as sensitive with an editable prompt backed by an explicit ‘always-redact’ and ‘never-redact’ list. Set how it runs:

• Suggest mode proposes redactions such as blurring a face in a research recording or muting a name in a sales call
• On mode applies all redactions automatically
• Off disables it entirely

Custom data retention works at both the workspace and project level, with project-level rules taking precedence when they’re set. Define how long data lives before it expires, and anyone affected by an upcoming deletion gets notified seven days in advance.

Version history shows exactly who changed what in a Doc and when, and lets a team roll back to any previous version, so shared context stays accountable as multiple people edit it.

Underpinning all this is ISO 42001 certification, the international standard for responsible AI management. It sits alongside Dovetail’s existing ISO 27001 certification and SOC 2 Type II compliance, giving legal and compliance teams a globally recognized bar to sign off against.

One playbook, every team

The bigger an organization gets, the harder—yet more essential—consistency is to maintain. Our new standardization capabilities have one project structure and one set of processes applied the same way across every team, so everyone in your organization’s working from the same playbook.

A custom home shapes what a team sees the moment they land in Dovetail, so the first thing anyone sees is the work that matters to their team, not a generic dashboard.

Project templates and Doc templates give teams a consistent starting point instead of a blank page, so a new research project or a new synthesis Doc follows the same structure the rest of the organization already uses.

Global fields and global tags keep the taxonomy consistent across the entire workspace. When every team tags a churn risk or a feature request the same way, that feedback rolls up into Channels and Agents cleanly, instead of fragmenting into a dozen slightly different labels that never quite match.

The enterprise-level security layer for your whole organization

Together, these capabilities are what make it safe to turn Dovetail on for everyone in your organization.

Context engineering means everyone works from the same clean, structured foundation—continuously organized so the answers given are always accurate. Governance gives you control over who can access what, with permissions, audit trails, and data handling that hold up to scrutiny from security and legal. Standardization ties it together in one consistent structure across teams and tools so everyone’s working from the same playbook.

Chat with our team to see how our enterprise-level security and governance features bring Dovetail to your entire organization safely.