fastn 2025 Year in Review: Building the Infrastructure for Production AI Agents

2025 was a breakthrough year for fastn. We shipped major platform updates, expanded our connector ecosystem, earned industry recognition, and completely refreshed our brand and product experience. As AI agents moved from experimentation to production deployments, we focused on building the infrastructure teams actually need to get their agents into the real world.

Before we look ahead to 2026, here is what we accomplished this year.

1. Bold Brand and Product Refresh

We started 2025 with a complete brand and website overhaul. fastn has evolved significantly since our early days, and it was time for our identity to reflect that growth.

The refresh touched everything: new visual design, restructured navigation, clearer product messaging, and a modernised user interface across the entire platform. We wanted teams evaluating fastn to immediately understand what we offer and how we can help them ship faster.

Beyond aesthetics, we redesigned the core product experience. The Connect UI received a complete makeover with clearer connector capabilities and streamlined workflows. We simplified onboarding so developers can go from sign-up to working integration in minutes, not hours. Every interaction was reconsidered with one question in mind: does this help our users move faster?

Our philosophy has always centred on making complex integrations simple. The refreshed brand now communicates that promise at first glance.

2. Major Features Shipped

We shipped four major releases throughout 2025, each focused on making fastn more powerful and easier to use.

September: Simplified Setup and Universal Embed (v1.0.0)

Our September release focused on removing friction from the integration process.

• •Modernised Connect UI: we rebuilt the connector interface from scratch. Each connector now includes concise capability descriptions, making it faster to evaluate which tools fit your needs.
• •Universal Embed Flow: a single, consistent process to integrate fastn into any client application. No more platform-specific setup guides, one flow works everywhere.
• •Sandbox Mode: teams can now evaluate fastn without creating an agent first. Start with just your LLM API key and server URL, test connectors, and see results before committing to a full implementation.
• •Updated Embed documentation: we added prerequisites and step-by-step instructions to the Embed section, creating a consistent setup path regardless of which platform you are building on.

November: Agent Creation and Data Sync (v1.0.1)

November brought automation capabilities that keep your agents current without manual intervention.

• •Automated Data Sync: configure custom sync schedules so your agent context stays fresh. Detailed execution logs make it easy to monitor sync status and troubleshoot issues.
• •Agent Creation Support: we streamlined the process of connecting new AI agents to fastn. The new workflow reduces setup time and guides you through configuration step by step.

November: Multi-Gateway Support and Schema Filtering (v1.0.2)

Later in November, we shipped features for teams running complex deployments.

• •Multi-Gateway Support: enterprise teams can run multiple gateways within a single workspace. Use separate gateways for different environments, use cases, or customer segments, all managed from one place.
• •Input and output schema filtering: granular control over what data flows through your agent connections. Filter schemas at the field level so agents only access the information they need.
• •Custom fastn workflow tools: extend automation beyond standard connector actions by building custom workflow tools that fit with the rest of your fastn setup.
• •Documentation updates: expanded guides covering the new features with practical examples and best practices.

December: Database Connectors and Schema Customisation (v1.0.3)

We closed out the year with our biggest expansion of connector capabilities.

• •Database connectors: direct connections between AI agents and relational databases. Connect to MySQL, PostgreSQL, SQL Server, and others without building custom middleware.
• •Rich schema customisation: fine-grained control over which fields and operations your agents can access, per connector.
• •Refreshed interface: visual updates across the platform that improve navigation and make common tasks faster.

3. Highlights and Awards

fastn was named the Top AI-Powered Embedded Integration Infrastructure Platform 2025 by CIOReview. The recognition came after an evaluation by a panel of C-level executives, industry thought leaders, and CIOReview's editorial board, and it reflects the trust our customers and industry peers have placed in the platform.

As the MCP ecosystem grew throughout 2025, fastn earned recognition across the leading directories where developers discover MCP tools and servers. We were featured in MCPulse, Archestra.io, and MCP.so, helping more teams find us as they build their AI agent infrastructure.

We also saw strong growth in our developer community. More teams than ever are building production AI agents on fastn, and their feedback shaped the roadmap directly. Every major feature shipped this year came from conversations with developers building real applications.

4. New Connectors Published

Our connector library expanded significantly in 2025, growing to over 500 pre-built integrations covering the tools teams rely on every day.

Highlights from this year's additions include Snowflake, Figma, Anthropic Claude, Gemini, LinkedIn, Airtable, ClickUp, Zendesk, and Jira, alongside dozens more across productivity, data, CRM, and developer tool categories.

We also launched database connectors as a new category, enabling direct connections between AI agents and relational databases like MySQL, PostgreSQL, and SQL Server without custom middleware.

Beyond adding new connectors, we improved the existing library with clearer capability descriptions, consistent action naming, better error messages, and expanded documentation for each integration.

Looking Ahead to 2026

2025 laid the foundation. In 2026 we are building on it with deeper capabilities, broader integrations, and continued focus on making production AI deployments straightforward.

We will share more detail on the 2026 roadmap in the coming weeks. For now, we are grateful to every developer, customer, and team member who made this year possible.