Tray.ai
Also known as: Tray.io, Tray
Enterprise integration and automation platform with an agent layer, offering more than 700 connectors, visual workflows, agents built without code and grounded in a synced knowledge base, and an MCP gateway that governs which tools agents may call.
Tray.ai, formerly Tray.io, is an enterprise orchestration platform for data and AI. One governed foundation carries integration, automation and the agents built on top of them. Workflows are built in a visual builder that needs little code, with conditional logic, branching, looping, retries and timeouts, drawing on a library of more than 700 managed connectors, and the same connectors are exposed to AI clients as MCP tools.
Merlin Agent Builder is where agents are assembled, from a scope that sets goals and limits, data sources, workflows as tools and a model. Connecting a data source such as Google Drive, Gmail, Confluence or SharePoint syncs the documents into a vector table the agent searches through a knowledge tool included by default, so answers are grounded in the customer's own material without live calls per query.
Agents can start from preconfigured accelerators, a Knowledge agent, an ITSM agent or a Support ticket agent, run on Tray's native model or on the customer's own OpenAI or AWS Bedrock keys, and reach people through Slack and Microsoft Teams as well as the API. A built in test chat shows each reply with the tools considered, whether the knowledge base was searched, processing time and token usage.
Agents keep short and long term memory, so they can refer back to a user's earlier conversations. A person stays in the path. Workflows take an approval step routed to Slack, email or a custom form, and agent guardrails pause for a human when confidence is low or an action falls outside the agent's scope. Tray's library also offers an evaluation template that scores agent responses with an LLM judge, so a team can compare versions of an agent over time.
Agent Gateway handles governance. Tray workflows become composite MCP tools and connector operations become individual tools, with an access allowlist, RBAC, rate limits, OAuth2 or token authentication for the calling client, and a choice between shared service credentials and the end user's own.
Runs are logged step by step, streamed to the customer's own endpoint, and maskable or switchable off where data minimization matters. Instances are hosted in the US, EU or APAC so data stays in region, with an on premises agent and AWS PrivateLink for systems behind the firewall. Compliance covers SOC 1 and SOC 2 Type 2, HIPAA, GDPR and CCPA. Pricing is quoted by sales across Pro, Team and Enterprise, metered in Tasks, with a free trial.
Vendor details
Canonical URL
https://tray.ai
Category
Agent infrastructure
Subcategory
Enterprise iPaaS + AI agent orchestration / MCP governance
Funding status
Independent; an established enterprise integration vendor, formerly Tray.io.
Company status
independent
Use cases & customers
Primary use cases
Target customers
Deployment options
Integrations
More than 700 managed connectors, including Salesforce, Workday, Snowflake, Jira, NetSuite and ServiceNow, are each governable through the same RBAC and publishable as MCP tools. Agent Gateway provides an MCP registry and governance layer. It covers composite MCP tools, tools backed by connectors and MCP consumption, with audit trails that can stream to Datadog and Splunk. Agents deploy to Slack, Teams, chatbots or APIs.
In practice
Your team spins up AI agents faster than IT can govern them, and shadow MCP servers multiply. Tray's Agent Gateway is an MCP registry that lets IT build, version, and publish governed MCP tools.
An agent is only useful if it can reach your stack. Tray connects through more than 700 managed connectors, so agents act across Salesforce, Workday, Snowflake, and the rest.
Ops leads need to build agents without waiting on engineering. Tray's Merlin Agent Builder lets them build without code, with guardrails, PII protection and a human in the loop built in.
Sources & related URLs
Research sources
Agentic Index coverage score
13.0 / 14 capabilities · 93%
| Integrations & Tool Calling | Full |
|---|---|
|
Agent Gateway exposes Tray workflows as composite tools and individual operations from 700+ connectors as tools over MCP, with tools authenticating by shared service credentials or the end user's own, an access allowlist, OAuth2 or API token for the calling client, RBAC and rate limits; agents built in Merlin Agent Builder take actions by running Tray workflows as tools. Sourcetray.ai/documentation/platform/artificial-intelligence/agent-gateway/overviewread 2026-09-22 |
|
| Workflow Orchestration | Full |
|
Tray is a workflow platform with an agent layer on top: workflows sequence steps with conditional logic, branching and looping, retries and timeouts, and an agent project combines a scope, data sources, workflows as tools and a model, so autonomous agent steps run alongside deterministic workflow nodes. Composite tools package multi-step business logic as a single callable tool. Sourcetray.ai/documentation/platform/artificial-intelligence/agent-builder/overviewread 2026-09-22 |
|
| Knowledge Grounding & RAG | Full |
|
Adding a data source in Merlin Agent Builder connects a system (Google Drive, Gmail, Confluence, SharePoint or a custom connector), syncs and processes the documents and stores the content in a vector table that supports semantic search, which the agent queries through a Knowledge Search tool included by default. New documents join that table without any retraining, and it persists and holds far more than fits in the model's context window. Sourcetray.ai/documentation/platform/artificial-intelligence/agent-builder/data-sourcesread 2026-09-22 |
|
| Human Oversight & Guardrails | Full |
|
In workflows, an approval step can be inserted anywhere and routed to Slack, email or a custom form, and the run pauses and resumes automatically once the decision is made, and agents built in Merlin Agent Builder act by running those workflows as tools. Guardrails also pause an agent when confidence is low or when an action falls outside its defined scope, and surface the decision to a person before it continues, and Merlin Agent Builder includes approval workflows for sensitive actions. Around that sit an Agent Gateway access allowlist, RBAC, rate limits, agent scope, log masking and an execution kill switch. Sourcetray.ai/glossary/human-in-the-loopread 2026-09-25 |
|
| Security, Identity & Governance | Full |
|
Compliance covers SOC 1 and SOC 2 Type 2, HIPAA, GDPR and CCPA, and Tray says every agent action, MCP tool call and workflow run is traceable to a user. Access controls cover organizations and workspaces, user roles, two factor settings, an access allowlist and RBAC on Agent Gateway, log masking to hide a step's inputs and outputs, and administration of which Merlin AI features are enabled. Sourcetray.ai/platformread 2026-09-22 |
|
| Observability & Auditability | Full |
|
Every workflow run produces debug logs showing each step's input and output, agents have their own logs and debugging view, Agent Gateway has an observability and monitoring page for tool execution, and Tray streams workflow logs to an external endpoint for the customer's own tooling, with log masking, disabling of log storage and anomaly detection as controls. Every agent action, MCP tool call and workflow run is traceable to a user. Sourcetray.ai/documentation/platform/enterprise-core/logs-debugging/log-streamingread 2026-09-22 |
|
| Memory & State Persistence | Full |
|
Merlin Agent Builder includes short and long term memory, so agents keep track of session history and refer back to a user's prior conversations automatically, with a sliding context window keeping each exchange focused. It is built for the case where an agent would otherwise forget a user's earlier issue every time that user returns, so memory carries across conversations. Knowledge synced into the agent's vector table is kept separately from memory. Sourcetray.ai/company/news/merlin-agent-builder-user-adoption-releaseread 2026-09-25 |
|
| Deployment & Data Residency | Full |
|
A Tray instance is hosted in a data region the customer picks, US, EU or APAC, so business data does not leave the region it originated in, with API base URLs for each region and setups across several regions for multinational or embedded use. For systems behind the firewall, Tray offers an on premises agent and AWS PrivateLink connectivity. Sourcetray.ai/documentation/platform/enterprise-core/organisation-management/regional-hostingread 2026-09-22 |
|
| Prebuilt Agents, Templates & Packs | Full |
|
Buyers start from preconfigured agent accelerators, each with a predefined scope for its use case: a Knowledge agent, an ITSM agent and a Support ticket agent, alongside prebuilt tools and workflow templates that can be customized. Sourcetray.ai/documentation/platform/artificial-intelligence/agent-builder/configurationread 2026-09-22 |
|
| Triggers & Channel Coverage | Full |
|
Workflows start from connector triggers and webhook subscriptions, scheduled polling for new data and an alerting trigger, and agents are reachable through interaction channels in Slack and Microsoft Teams as well as the API. Sourcetray.ai/documentation/platform/artificial-intelligence/agent-builder/interaction-channelsread 2026-09-22 |
|
| Model Flexibility & Routing | Full |
|
Each agent runs on Tray's native default model or on the customer's own, connected through API keys from third party providers such as OpenAI or AWS Bedrock, with the provider and the specific model version chosen in the agent dashboard. Sourcetray.ai/documentation/platform/artificial-intelligence/agent-builder/configurationread 2026-09-22 |
|
| APIs, SDKs & MCP Extensibility | Full |
|
Tray publishes OpenAPI specifications for the Tray Platform and Tray Embedded APIs, a Connector Development Kit for building and testing custom connectors, developer tools including a connector tester and operations explorer, embedded auth dialogs for the customer's own end users, and Tray Headless MCP for connecting Cursor, Windsurf, Codex or Claude Code to the platform. Sourcetray.ai/documentation/platform/tray-headless/tray-headless-mcpread 2026-09-22 |
|
| Testing, Debugging & Optimization | Full |
|
Tray's library offers an AI evaluation framework, a template that uses an LLM as a judge to review each input prompt and agent response against set criteria, covering response quality, conversation flow, edge cases and behavioral consistency. The judge returns structured feedback and quantitative metrics tracked over time, so a team can compare versions of an agent and see whether a change improved it. A built in test chat shows each reply with the tools considered, whether the knowledge base was searched, processing time and token usage. Sourcetray.ai/documentation/library/template/6480885f-d884-469c-8c80-77a2f36a4b61-get-started-with-ai-evaluationsread 2026-09-25 |
|
| Browser & Computer Use | Not documented |
|
Tray acts through connectors, workflows and MCP tools that call APIs. It does not describe a browser, desktop or remote computer session it drives for an agent, or a headless fetch or third party browsing engine of its own. Sourcetray.ai/documentation/platform/artificial-intelligence/agent-gateway/overviewread 2026-09-22 |
|
The Agentic Index coverage score grades every vendor Full, Partial or Not documented against the same 14 buyer facing capabilities, from public evidence only. Each capability links to how all vendors in the index score on it. How this evidence is graded
Product updates
Tray.ai introduced Tray Helix, a governed runtime environment designed for AI-built applications. The release features a centralized credential broker that allows apps to reference authentication aliases, ensuring that raw secrets never touch application code or AI prompts.
Bears on: Security / enterprise
View sourceTray.ai made MCP dynamic authentication generally available. MCP tools can now run using end-user credentials instead of shared service accounts, with allowlists, authentication management, and a usage monitor.
Bears on: MCP / tool calling / API
View sourceTray Headless MCP added region-specific endpoints for US, EU, and APAC workspaces.
Bears on: Deployment / data residency
View sourcePricing
Custom quote for every tier (Pro, Team and Enterprise). A free trial is arranged through sales.
Usage, metered in Tasks across integration, automation, MCP and agents. Task volume is quoted per customer.
Included quota
Pro covers up to 3 workspaces with 7 day insights retention. Team covers 20 workspaces with 30 day insights retention, enhanced insights, 24/7 support, and dedicated Slack. Enterprise covers unlimited workspaces with 180 day insights retention, regional hosting, log streaming, on premise connectivity, and embedded white label options. All plans include business hours support and a free trial. Task volume is quoted per customer.
What is public
The three tiers are published with what separates them (workspaces, insights, support, regional hosting), along with the 700+ connectors and connectors as MCP tools in every plan and the metering in Tasks. No rate is published, in any unit.
Billing mechanics
A usage based subscription metered in Tasks across three tiers, quoted by sales on expected usage and the features required. Multi step automations, real time triggers and retries all consume Tasks. Agent Builder and Agent Gateway are enabled by Tray's account team, not self serve.
Cost watchouts
Task consumption is driven by automation complexity and trigger frequency, so real usage can exceed early estimates, and Agent Builder and Agent Gateway must be switched on by Tray's account team. Regional hosting is an add on per region on Team and standard only on Enterprise. Support SLAs, insights retention and workspace count all move with the tier.
Variable cost rationale
Tray meters Tasks. Multi step automations, real time triggers and retries all consume them, and consumption grows with workflow complexity, trigger frequency, data volume, connector count and retry logic, so heavy automation and agent activity drive the bill up quickly. Add ons such as HIPAA, extended retention and the separately licensed agent product add further cost. The quoted plan sets the task volume, and going past it means a higher tier or a larger allotment agreed with sales. No per Task rate is published.
Additional watchouts
Pricing is quote only with no public dollar figures. The agent builder is a separate line item from core iPaaS, HIPAA is an add on, and task based billing means cost scales with automation volume.
Overage / add-ons
Task usage is set by the quoted plan. Exceeding it means moving to a higher tier or agreeing a larger task allotment with sales.
Sales call required
Mixed (some tiers require a call)
Free / trial
Free trial arranged through sales, and no free plan
Lowest paid plan
Not published
Commercial notes
A single AI native platform spanning integration, automation, MCP governance and agents, positioned to replace fragmented legacy iPaaS stacks. Tray says customers ship faster and spend materially less than on legacy iPaaS. Under the usage based, quoted model, the bill depends on task volume.
Key ambiguities
No dollar figure is published, and every plan routes to a custom quote based on usage and features. What a Task costs, and how agent usage is metered against Tasks, are set by sales.
Support SLA / resale
Business hours support (Tray Advantage) on all plans, and 24/7/365 support, enhanced SLAs and dedicated Slack channels (Tray Advantage Plus) on Team and Enterprise. Enterprise adds an embedded white label bundle.
Missing data
Dollar rates for each tier, per task pricing, the separate Merlin Agent Builder license cost, and add on pricing for HIPAA and extended retention are not public.
Related vendors
- AgentOps — Agent observability and debugging platform: open source SDKs trace…
- AgentPhone — AgentPhone gives the AI agents a team builds their own US or…
- Agno — Python agent framework and AgentOS runtime (formerly Phidata) for…
- AIsa — Resource and payment gateway for AI agents, with one key to 110+…
- Albato — Albato is a no code integration platform that sells an embedded,…
- AlphaBitCore — AI control plane for regulated financial firms: one gateway enforces…
Alternatives to Tray.ai
The closest documented capability profiles to Tray.ai among agent infrastructure platforms tracked by Agentic Index, ordered by similarity on the same 14 point evidence the rankings use. No vendor pays for placement.
- Agno13.0 / 14Matches Tray.ai across all 14 documented capabilities
- Kestra12.5 / 14A lighter documented profile than Tray.ai
- Haystack12.0 / 14A lighter documented profile than Tray.ai
- LlamaIndex12.0 / 14A lighter documented profile than Tray.ai
- Mastra14.0 / 14Adds documented Browser & Computer Use
- xpander.ai12.0 / 14A lighter documented profile than Tray.ai
Similarity is computed from each vendor's Agentic Index coverage score evidence, axis by axis, not from the totals. How this evidence is graded