AI that makes it past the pilot.

Central teams can control user roles, provider integrations, model access, API tokens and tool permissions. Workflows can require human approval or user authentication before sensitive actions continue, and every run retains a detailed trace.

The platform runs as independent regional deployments in Ireland, the United States and Canada, each with its own database and infrastructure. Gecko does not train AI models on institutional data. When you bring your own provider keys, model inference runs under the data-processing agreement your institution already holds with that provider.

Your existing Gecko Engage products keep working exactly as they do today—and the platform connects to them, so agents can act across Gecko Engage contacts, events, forms and campaigns through a hosted integration. When you’re ready, we’ll help you evolve existing chatbot use cases into governed platform agents on your own timeline.

Gecko provides a central prompt library that teams can organise into folders and manage through role-based read and write permissions. Saved prompts support dynamic variables, dates and conditional logic, and can be reused by agents and workflow AI steps instead of being copied into every implementation.

Yes. Gecko can ground agents in websites and managed knowledge, make your student information system agent-ready through Ellucian Ethos, connect Salesforce, and work with SharePoint, Google Drive, OneDrive, Dropbox and S3. Beyond that, agents act through integrations, APIs, webhooks and MCP—connect existing MCP servers or build your own to give agents controlled access to proprietary systems.

Choose a floating chatbot, a minimal inline conversation or an AI search experience, then add the generated script tag to an approved website. Each interface can use tenant-specific colours, title text and imagery. Shadow DOM style isolation prevents the host website’s CSS from changing the interface, while keyboard, screen reader and reduced-motion patterns support accessibility review.

No. The platform supports OpenAI, Anthropic, Amazon Bedrock, Azure OpenAI, Google and OpenRouter, plus any OpenAI-compatible endpoint your institution approves. You can bring institution-owned API keys or use Gecko-managed access, and swap the model behind a live agent without rebuilding it. Tools use the open MCP standard and workflows export as portable JSON—so your investment travels with you.

Start with a bounded outcome and baseline the current service. Unified observability then shows run volumes, status, duration, token usage, tool calls and errors across agents and workflows. Combine that evidence with enquiry volumes, handling time and cost data to build a value case your finance and service leaders can test.

With Gecko’s managed implementation, a fully operational agent is deployed in less than a week on average. We help define the use case, connect knowledge and tools, establish controls, test behaviour and support go-live.

Yes. Single sign-on—including SAML—is provided through Gecko’s identity service, and the console is SSO-only in production. Role-based access control separates read, write and run permissions across 26 resource types, and API tokens are scoped, expiring, stored only as hashes and can never exceed the permissions of the person who created them.

The platform is hosted on AWS with SOC 2 and ISO 27001 assurance, deployed in independent regional stacks in Ireland, the United States and Canada. Credentials and secrets are AES-256-GCM encrypted, and a VPAT supports accessibility review—with the administrative interface available in English, Welsh and Spanish. A full security pack is available on request.

Start with one bounded process where demand is high, knowledge is trusted and the outcome is measurable. Gecko can help shape that first use case, establish the guardrails and turn what works into a repeatable pattern for other teams.