Token Efficiency
Only the context a task needs, on the model it needs, unmetered inside your tenant.
Learn moreMarketplace
AI and Tools Marketplace
Choose what your teams use. Approve each one.
- In-tenant models, or the OpenAI and Claude APIs under policy
- Jira, GitHub, GitLab, Azure DevOps, SonarQube, Slack, Teams
- Admin approval before anyone can use it
What is the SyncupHUB AI and Tools Marketplace?
The SyncupHUB marketplace is where an administrator chooses which AI models and delivery tools the organisation may use. Models can run inside your own tenant or come through an external provider’s API under policy, such as the OpenAI API behind ChatGPT, Anthropic’s Claude API or another AI provider, and tools connect Jira, GitHub, GitLab, Azure DevOps, SonarQube, Slack and Microsoft Teams. Nothing reaches your teams until it is approved.
How it works
Browse the catalogue
Models and tools are listed with what they are for and where they run, in your tenant or through a provider such as OpenAI or Anthropic, rather than one model hard-wired into every feature.
Approve by policy
An administrator approves each provider and tool and decides which kind of work it may be used for.
Keep it in your tenant
Models can run inside your own tenant, next to the delivery data they read. Reaching an outside provider is a choice, not a default.
Connect the tools
Tools connect with the field and hierarchy mapping you define, and the Integration Agent checks whether each connection is behaving.
What’s included
- Local model services in your tenant
- OpenAI API, the GPT models behind ChatGPT, under policy
- Anthropic Claude API under policy
- Other AI providers’ APIs, approved the same way
- Jira two-way work synchronisation
- GitHub, GitLab and Azure DevOps engineering evidence
- SonarQube code quality
- Slack and Microsoft Teams notifications and approvals
- SSO and SAML sign-in with SCIM provisioning
- S3-compatible attachment storage
What it needs
Stated plainly, before a demo does.
- What each connector covers depends on the connector.
- Integration-dependent features need their connection configured first.
- External AI providers are used only if you approve them.
- An external AI provider bills for its own usage, on its own terms.
Questions about AI and Tools Marketplace
Does SyncupHUB work with the ChatGPT API and the Claude API?
Yes. An administrator can approve the OpenAI API, which serves the GPT models behind ChatGPT, Anthropic’s Claude API, or another AI provider’s API. Agents and AI features then use only the providers approved for that kind of work, and the approval is recorded.
Can we use more than one AI provider?
Yes. Each provider is approved separately and for the kinds of work it may be used for, so an organisation can use different models for different work, or keep some work on models inside its own tenant.
Why a marketplace rather than one built-in AI model?
Because no single model is right for every task or every data boundary. A catalogue of providers lets an administrator decide which ones your organisation may use and for what kind of work, and agents and AI features use only the providers that have been approved.
Can we run AI models inside our own tenant?
Yes. Models can run in your own tenant, self-hosted or in your cloud tenant, next to the delivery data they read, so nothing has to cross a boundary to answer a question. Using an external provider is a separate decision your administrator makes.
Which tools does SyncupHUB connect to?
Jira for two-way work synchronisation; GitHub, GitLab and Azure DevOps for engineering evidence; SonarQube for code quality; Slack and Microsoft Teams for notifications and approvals; enterprise identity with SSO, SAML and SCIM; and S3-compatible storage for attachments.
Who decides which models and tools teams can use?
Your administrator. Each provider and tool is approved before anyone can use it, the approval is recorded, and AI access and provider controls sit with the organisation’s security settings.
OpenAI, ChatGPT and GPT are trademarks of OpenAI; Claude and Anthropic are trademarks of Anthropic. They are named to describe compatibility; no endorsement is implied.
Related features
AI Agents for Every Role
Seventeen persona agents and 59 sub-agents, all reading the same delivery model.
Learn moreRequirement to Code Traceability
Follow a requirement to the code that shipped it, and recover requirements from code that exists.
Learn moreFurther reading
See AI and Tools Marketplace on your own delivery
A 30-minute walkthrough on your delivery model, not a generic demo.