Why This Job is Featured on The SaaS Jobs
This Product Manager seat sits at the intersection of SaaS platform design and the emerging “AI agent” layer now being added to established cloud products. The listing signals a mature, widely adopted SaaS environment where AI is becoming a primary interaction surface, which makes the work less about novelty features and more about integrating autonomy, trust, and workflows into a product used for real operations.
From a SaaS career perspective, the role builds durable product muscles around adoption mechanics, measurement, and iteration in areas where outcomes are harder to define than traditional feature delivery. Owning an end to end experience that spans onboarding from existing customer data, workflow creation, and ongoing improvement is a strong proxy for learning how SaaS products compound value over time. The emphasis on evaluation frameworks and architecture tradeoffs also maps well to the broader shift toward AI enabled product strategy across B2B software.
This is best suited to product leaders who prefer working close to technical constraints while still being accountable for user facing quality. It will fit someone comfortable making decisions amid evolving patterns, collaborating across multiple product surfaces, and treating metrics as a design tool rather than a reporting artifact.
The section above is editorial commentary from The SaaS Jobs, provided to help SaaS professionals understand the role in a broader industry context.
Job Description
In the last six months AI agents have crossed a meaningful threshold, evolving from turn-by-turn assistance to autonomous systems that reason, plan, and execute complex work without intervention. Intelligence is no longer the bottleneck. Connecting it to real organizational data and workflows is the hard part, and the companies that build the right infrastructure now, while the stack is still being defined, will be difficult to displace.
Airtable is building the platform for human-agent collaboration: a shared operational surface where humans and agents work together on the same data, with the same visibility and the same rules. Omni is the agent at the center of that vision. It lets anyone build apps, automate workflows, and get answers from their operational data through natural language. It's the primary surface through which customers experience AI in Airtable, and the vehicle through which we make good on the promise that Airtable becomes more valuable as AI proliferates.
Think of Omni as a forward-deployed Airtable engineer in every base that understands your data and helps turn anyone into a builder.
This role owns the full Omni experience: the build loop (turning intent into working apps and workflows) and the intelligence layer (analysis, Q&A, and synthesis over operational data). You'll be responsible for making Omni the best way for business users to build and operate AI-powered apps and workflows.
Why this matters now:
- AI is reshaping how software gets built and used. Tools like Lovable, Replit, and Cursor have shown that AI can build real applications, but none of them solve for structured data, collaboration, or enterprise trust. That's the gap Omni is built to own.
- The scope is broad. This role spans the full AI experience in Airtable, from building apps and workflows to configuring agents to analyzing data. You own the primary way customers interact with AI across the product.
- Enterprise customers are ahead of us. Our largest accounts are already running AI-powered workflows and pushing us on what's next. The demand is specific, urgent, and shapes our roadmap directly.
- Massive distribution. Half the Fortune 100 are customers and we have hundreds of thousands of monthly self-serve signups. Omni improvements reach enormous scale immediately.
What you'll do
- Own the end-to-end Omni experience. From the moment a user describes what they need, through app creation, workflow configuration, and ongoing iteration. Omni should feel like an expert consultant, not a chatbot.
- Push the frontier of what Omni can do. Partner closely with engineering to evolve Omni's agent architecture, define how it reasons about and executes multi-step plans, and ensure it compounds with model improvements rather than requiring piecemeal capability additions. As models get smarter, Omni should get smarter automatically.
- Make app building dramatically better. Raise the quality ceiling on what Omni can build: richer interfaces, more sophisticated automations, better data modeling. We should be the best way to build collaborative operational apps.
- Make it easy to start from real data. Most people don't build from a blank canvas. They have messy spreadsheets, file exports, or data living in a warehouse. You'll own the experience of going from "here's what I have" to "here's a working app," making Omni the fastest path from raw data to something useful.
- Define and drive adoption metrics. Establish north star metrics for Omni that capture whether we're making people more successful, build a measurement framework, and use it to drive roadmap prioritization.
- Collaborate across teams. Work closely with the teams that own agents, MCP, integrations, enterprise, and self-serve to ensure Omni is coherent across all surfaces.
Who you are
- Experience shipping AI agent or copilot products. You've built products where an LLM reasons, plans, and takes action, not just generates text. You understand agent loops, tool use, context management, and the difference between a demo and a production-quality agent experience.
- 8+ years of product management experience with a track record of shipping products users love, ideally including AI-powered features at scale.
- Deep AI technical fluency. You understand how LLMs work at a technical level, not just as a user. You can evaluate tradeoffs between model capabilities, have opinions on agent architectures, evals, prompt design, and know what it means to build so that your product gets better as models get better.
- Strong product taste for AI-native building experiences. You have a refined sense for what makes an AI building experience feel magical vs. frustrating. You've thought about the spectrum from fully autonomous (agent does it all) to highly collaborative (agent and human iterate together) and know when each is appropriate.
- Builder mindset. You use AI tools (Claude Code, Cursor, Lovable, etc.) to build and prototype, not just spec. You can evaluate Omni's quality by using it, and by using competitors, with the critical eye of someone who builds things themselves.
- Comfort with ambiguity and architectural decisions. Omni is evolving fast. You need to be energized by foundational questions (what should Omni's planning loop look like? how do we handle failure and recovery? where do we draw the line between agent autonomy and human control?) and not just feature sequencing.
- Analytical rigor. You can define success metrics for inherently fuzzy AI outcomes, build measurement frameworks, and use data to distinguish real quality improvements from noise.
