How Smartsheet built a remote MCP server on AWS

2026-07-17 · Source: Artificial Intelligence · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Cloud Computing & IT Infrastructure · Depth: Advanced, long

Summary

Smartsheet developed a remote Model Context Protocol (MCP) server on AWS, enabling AI agents and assistants like Amazon Quick and Claude Desktop to directly access its enterprise work management platform. This architecture, built on AWS Fargate for Amazon ECS, Amazon Kinesis, Amazon S3, Amazon Bedrock, and Amazon Neptune, facilitates natural language interaction and autonomous AI workflows. Key optimizations include progressive disclosure for token budgeting, schema-driven tool contracts using Pydantic to prevent hallucination, and a proprietary serialization format that reduced token consumption by 35-47% and saved over 3 billion tokens. The system incorporates robust deployment strategies with ECS Auto Scaling and CI/CD circuit breakers, layered security using AWS WAF and AWS Shield, and comprehensive observability via OpenTelemetry, Amazon OpenSearch Service, and Datadog. This infrastructure supports bursty AI traffic and ensures governance through tiered access control and audit trails.

Key takeaway

For AI Architects and MLOps Engineers designing enterprise AI agent integrations, prioritize token optimization and burst-aware scaling from the outset. Implement layered security and governance directly into your AI tool frameworks, ensuring access control and audit trails are default. Validate non-deterministic AI workflows with end-to-end tests that include the LLM in the loop, ensuring reliable and cost-effective agent operations.

Key insights

Smartsheet's AWS-hosted MCP server provides AI agents structured, token-optimized access to enterprise data, ensuring reliability and cost efficiency.

Principles

• AI agent infrastructure requires burst-aware scaling.
• Token optimization is critical for LLM cost and performance.
• Governance and security must be built into AI tool frameworks.

Method

Smartsheet's MCP server architecture integrates AWS Fargate, Kinesis, S3, Bedrock, and Neptune, applying progressive disclosure, Pydantic-driven schemas, and proprietary serialization for AI-optimized data access.

In practice

• Implement layered rate limiting for AI agent traffic.
• Use OpenTelemetry for agent-first identity and tracing.
• Validate tool behavior with LLM-in-the-loop tests.

Topics

• Model Context Protocol
• AWS Architecture
• AI Agent Orchestration
• Token Optimization
• Enterprise AI Governance
• LLM Testing

Best for: CTO, VP of Engineering/Data, Director of AI/ML, AI Engineer, MLOps Engineer, AI Architect

Related on AIssential

• Article: Building a Secure MCP Server on AWS for a Million-Company B2B Platform — Treating MCP servers as first-class interfaces with narrow tool contracts enhances security and maintainability for LLM integrations.
• Extending MCP support for Amazon Bedrock AgentCore Gateway — AgentCore Gateway centralizes MCP server governance, enabling scalable, secure, and stateful enterprise agent workflows.
• Building agents that reach production systems with MCP — MCP standardizes agent-to-system integration, enabling scalable, portable, and context-efficient connections for cloud-based AI agents.
• MCP Dev Summit [Day 2] ft AWS, Docker, & Datadog — MCP is evolving to support real-time, interactive, and secure agent experiences, moving beyond simple request-response models.
• Building agentic AI applications with a modern data mesh strategy on AWS — Agentic AI requires multi-layered, deterministic governance across all data interaction steps, unlike single-checkpoint RAG.
• MCP as the API for AI‑Native Systems: Security, Orchestration, and Scale — MCP is becoming the API for AI-native applications, requiring secure, orchestrated, and scalable deployments.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Artificial Intelligence.