Senior Software Engineer, AI

Muck Rack is trusted by thousands of organizations around the world to understand how their brands show up across news, social media and AI-generated answers.

Our AI-powered platform connects the entire PR workflow in one place, helping communications teams manage media relationships, understand the conversations shaping their brands, take action and measure their impact. Combined with analyst-led Media Intelligence and advisory services, Muck Rack turns signals into insights and insights into action.

Founder-controlled and growing profitably and sustainably, we have the freedom to build forward-thinking products, put our customers first, and offer an inclusive culture that welcomes diverse perspectives.

Our Agents team at Muck Rack is building AI for PR and communications professionals. Our first agent helps customers understand their media coverage. Next, we’re building agents that can plan and carry out multi-step workflows.

You’ll take features from idea to production, working across our Python agent runtime, MCP server, and Django/Vue application. You’ll help decide how the technology works, ship it to real customers, and learn from how they use it.

You’ll help shape what we build next and how we build it. You’ll work with new models, solve real customer problems, and take your ideas from experiment to production. If you want to build agents people actually rely on, we’d like to meet you.

• Build agents that take action. Build tools, prompts, guardrails, and multi-turn workflows with memory. Help agents know when to act, ask a question, or hand off to a human
• Ship the full experience. Build in-product chat, streaming responses, permission-aware APIs, and our external MCP server
• Make quality measurable. Build evals, catch regressions, and use production feedback to improve quality, latency, and cost
• Help shape the platform. Make decisions about how we build, deploy, and run agents as the product grows

• 7+ years building production software
• Fluent Python skills, including async programming and experience with FastAPI, FastMCP, or Django
• Experience shipping LLM-powered features or agents to real customers including exposing tools via MCP
• A practical understanding of prompting, tool use, context management, and where models go wrong
• Experience measuring AI quality and using what you learn to improve it
• Strong API and tool design skills, including authentication and authorization
• Proactively incorporates AI tools into day-to-day work to improve productivity and accelerate delivery
• Working-hour overlap with US time zones (Eastern preferred)

• Agent frameworks such as Pydantic AI, Lang Graph, or the OpenAI Agents SDK
• MCP servers and LLM tooling
• AI evaluation tools and methods, including Deep Eval and LLM-as-judge
• Streaming UIs, server-sent events, and Vue.js
• Experience with LLM observability
• AI security, including prompt injection defenses and tenant isolation

Below you'll find an outline of the interview plan for this role. Please note that this is what we expect the process to look like; we may ask you for supplemental information or require an additional step before making a final decision.

• 30-minute interview with a member of our Talent Team
• A 1 hour zoom interview with the hiring manager
• Peer interviews with several team members
• 30-minute Executive Interview
• 30-minute Bar Raiser

This role requires up to 10% travel for team collaboration, customer engagements, and company events. As part of our commitment to building strong connections across our fully distributed team, attendance at our annual company offsite…