apidays Munich 2025 - Integrate Your APIs into the New AI Marketplace, Senthil Avinash (Postman)

apidays Munich 2025 - Integrate Your APIs into the New AI Marketplace, Senthil Avinash (Postman)

• 1.
  All rights reservedby Postman Inc Senthil Avinash Strategic Solutions Engineer, Postman Integrate your APIs into the new AI Marketplace
• 2.
  ● Strategic SolutionsEngineer driving API- first innovation at Postman ● AWS Certified AI Practitioner ● Previously at Salesforce & Guidewire, enabling financial services and insurers with API-led transformation ● Volunteer photographer at Ronald McDonald House Charities of Denver ● Based in Denver, CO – enjoys hiking and biking in the Rockies About me GET api.getpostman.com/me
• 3.
  Large Language Models(LLMs) are becoming active agents, needing access to tools and data to perform complex tasks. This will create a new marketplace for your services. AI is evolving The new AI Marketplace
• 4.
  Integration Challenge The M×NProblem ● Each AI model needs unique connections to every service. ● This complex setup slows progress and costs more to maintain. ● Scaling AI is impossible without a simpler, unified solution
• 5.
  Solution Model Context Protocol(MCP) ● MCP is a standard protocol for AI to use tools and data. ● It simplifies integration reducing complexity to M+N. ● Provides a consistent way for LLMs to connect to various external resources
• 6.
  MCP Architecture How ItWorks ● MCP Hosts are AI apps where users interact. ● MCP Clients manage secure connections to Servers. ● MCP Servers expose capabilities from existing APIs.
• 7.
  MCP Flow Typical CommunicationFlow ● User Interaction: User submits a query to the Host. ● Host Processing & Connection: Host identifies required external capabilities and connect to Server(s). ● Capability Discovery & Invocation: Client queries Server for available tools and fetches capabilities to use. ● Server Execution and Result: Server performs an action and returns results to Client. Client send results to Host.
• 8.
  Tools => APIs Agentsneed tools. Those tools are your APIs…
• 9.
  ● APIs arenow for AI agents, not just developers. ● Design APIs as logical tools AI can easily use. ● APIs need machine-readable metadata for AI discovery. ● Measure success by AI Agents usage. AI Ready Preparing APIs for Agents
• 10.
  ● Centralize &Organize APIs ● Make APIs Discoverable ● Document and add Rich Types. ● Test and monitor APIs ● Share APIs with stakeholders API Standard Ready for MCP Consumption To make your API consumable by the Model Context Protocol (MCP), expose your endpoints in a way that MCP-compatible clients and AI assistants can discover, understand, and invoke them securely and efficiently.
• 11.
  ● Use OAuth2.0 / RBAC for secure access ● Encrypt data (TLS 1.2+ and at rest) ● Apply rate limiting, IP whitelisting, and input validation ● Use secure CI/CD with code scanning and reviews ● Enable MFA and manage credentials securely ● Follow least privilege access ● Stay alert to phishing; update passwords/software ● Review app permissions and data sharing Security Best Practices Developers & Users
• 12.
  ● MCP isbacked by Anthropic, Google, Microsoft and OpenAI. ● Many companies are rapidly adopting MCP. ● A large open-source community supports MCP's growth. ● MCP Catalog - public workspace of curated ready-to-use MCP servers. MCP Momentum A Growing Ecosystem
• 13.
  ● Supporting MCPbuilds AI-native platforms. ● MCP lets AI agents leverage your services. ● Ignoring MCP risks platform isolation and irrelevance. Strategic Decision Enabling the Future Enable the Future or Fall Behind NOW IS THE TIME TO ACT!!
• 14.
  Educate teams onMCP and API design changes. Evaluate existing APIs for MCP mapping. Experiment with MCP prototypes and tooling. Educate Evaluate Experiment Action Plan Leverage Context Aware API Platforms
• 15.
  Thank You!