Staff Software Engineer – Core AI Platform (MCP & Agent Infrastructure)

About this role

As a
Staff Software Engineer
on the
Core AI Platform
team, you will lead the design and development of the foundational platform that powers
Dojo AI agents
at Sumo Logic. You will own the architecture and implementation of a
robust, scalable MCP (Model Context Protocol)–based platform
that enables AI agents to securely access context, invoke tools
securely
, and interact with enterprise systems in real time.


In this role, you will build and operate the
core MCP server infrastructure
, including frameworks for hosting first‑party MCP servers, federating with external and third‑party MCP servers, and orchestrating agent interactions across distributed environments. You will design systems that allow agents to reason over
real‑time and historical context
, manage conversation state and memory, and reliably execute tool calls with strong guarantees around fault tolerance, retries, idempotency, and isolation.


You will also architect the
agent communication layer
, enabling seamless integrations with systems such as
Slack, Microsoft Teams, and other event‑driven communication platforms
, allowing agents to be invoked from messages, threads, and workflows. Your work will ensure secure handling of credentials, webhooks, and streaming events while supporting multi‑tenant execution at enterprise scale.


A key part of your responsibility will be building
deep observability and reliability
into the platform—providing visibility into MCP interactions, agent decision paths, and tool executions; enabling monitoring, alerting, and debugging of non‑deterministic agent behavior; and enforcing rate limits, quotas, and backpressure to ensure platform stability.


Responsibilities

• Architect MCP-first platforms

• Design scalable, fault-tolerant infrastructure for hosting and operating MCP servers
• Define standards for MCP server onboarding, versioning, and interoperability

• Build federated context systems

• Enable a...