Operations Engineer

Job description

**About Iru** Iru is the AI-powered security & IT platform used by the world’s fastest-growing companies to secure their users, apps, and devices. Built for the AI era, Iru unifies identity & access, endpoint security & management, and compliance automation—collapsing the stack and giving IT & security time and control back. Iru is backed by some of the smartest investors in tech—General Catalyst, Tiger Global, Felicis, Greycroft, and First Round Capital. In July 2024, Iru raised $100 million from General Catalyst, valuing the company at $850 million. Customers include Cursor, Vercel, Lovable, Replit, and Mercor, and Iru partners with industry leaders such as ServiceNow and AWS. Iru was named to Forbes’ America’s Best Startup Employers 2025 list for employee engagement and satisfaction. **The Opportunity** We're hiring an **Operations Engineer** — the first builder in Iru's Operations function, reporting to the Head of Operations. This role sits at the intersection of business operations and engineering. You'll need an unusual combination: enough operational judgment to know which problems are worth solving, and enough technical capability to go solve them yourself. Most people have one or the other. We're looking for someone who has both and wants a seat where both matter. Iru already has the people running the work. Accounting closes the books, FP&A runs planning, and there's analytics support across the business. This role doesn't take that work over. It exists to change the shape of it. The thesis, stated plainly: as Iru grows, operational load grows with it. The default answer is to add a person every time the load goes up. We think that's the wrong answer, and we'd rather prove it than assume it. Your job is to build the systems, integrations, and automations that give capacity back to the teams who have it today — so the work scales with the business without the headcount scaling in lockstep. You won't be measured on tickets closed or processes documented. You'll be measured on how much capacity you return. **Where we're starting.** The clearest near-term opportunities sit in G&A, and a few are already on the table: - **Collections.** Our AR process works, but it works because people are pushing it. We want it to work because the system is. - **ERP implementation.** We're moving to a new tool. You'll own how it connects to everything upstream and downstream. - **FP&A tooling.** We're selecting a planning platform. You'll help evaluate it, then own how it gets wired in. - **Data infrastructure.** Finance, people, and spend data lives in too many places and reconciles too slowly. You'll fix the plumbing. Those are examples, not the job. The same problem shows up in different clothing across T&E, People, Workplace, and planning. Where you go after the first wins is something you and the Head of Operations decide together, based on where the leverage actually is. ### What You'll Do ### Build - Design, build, and maintain integrations and automations across our internal stack so data moves between systems instead of through people - Write real code where the tooling runs out — scripts, services, jobs. When an off-the-shelf product doesn't do what we need, build the thing that does. - **Build internal applications people actually use.** A tool that replaces a spreadsheet. A form that replaces an email thread. An interface that replaces a manual queue. Internal software fails on adoption far more often than on code, so you'll build it with the people who'll use it and keep iterating after launch. - **Put AI where it earns its place.** Agents and model-driven steps inside real workflows — with evals, guardrails, and a human in the loop wherever judgment or money is involved. We're an AI company; our internal operations should look like it. - Own automation quality end to end: logging, error handling, monitoring, retries, secrets management, and clear ownership when something breaks - Build human approval steps into anything touching money, access, or the ledger — automation should remove the tedious parts of a control, never the control itself - Partner with Data and Engineering on pipelines, source-of-truth definitions, and reporting people actually trust ### Find the leverage - Sit with the people doing the work today, map what actually happens (not what the SOP says), and find the steps that shouldn't need a human - Redesign workflows alongside the people who own them — they own the process and the controls; you own the redesign and the build - Evaluate and select new tooling when buying beats building, and own how it lands - Baseline the before state, instrument the after, and report the delta honestly — including when something didn't pay off ### Our stack HubSpot, Stripe, Snowflake, Hex, Notion, and Campfire (incoming ERP), plus an FP&A platform we're selecting now. If you've done this work on Salesforce, NetSuite, or Workday, the systems are different but the job is the same — we care that you've connected things and automated them, not which vendor's logo was on the console. ### What You'll Bring - **2+ years building and scaling operations with a technical skew.** Business operations, product operations, revenue operations, business systems, analytics engineering, or a technical role that drifted toward the business. Titles vary a lot here; the shape doesn't. - **Something you built that runs without you.** This is the one requirement we won't flex on, so we'd rather be upfront: we'll ask you to walk us through a system you built — what it replaced, what broke, how you found out it broke, and whether it still runs now that you've moved on. A process you designed, or a project you scoped for someone else to build, won't get there. Rough and homemade is completely fine. It has to be a thing that runs. - **Python and SQL, hands-on.** You'll be reading and reconciling data across systems constantly and writing the code that moves it. You need to do that without waiting for someone else. - **Integration fundamentals.** REST APIs, webhooks, JSON, auth patterns — and the ability to debug a broken integration from logs and API responses rather than filing a ticket. - **You build with AI, not just alongside it.** Those are two different skills and we want both. You reach for coding agents by default and have shipped things you couldn't have shipped without them — and you've also built systems that *use* models: an agent that handles a workflow end to end, a model-driven step inside a pipeline, prompts you had to actually evaluate rather than eyeball. You know where models are reliable, where they aren't, and how you'd tell the difference in production rather than in a demo. - **You've built internal applications.** Not only scripts — things with a user on the other end. Retool, Streamlit, a small web app, an internal tool of any shape. You've watched someone use something you built, seen them get stuck, and gone back and fixed it. - **Enough business literacy to pick the right problems.** You don't need to be an accountant. You do need to understand how things like AR, AP, close, and forecasting work well enough to redesign them credibly — and to know when to defer to the people who own the controls. - **The ability to change other people's processes without alienating them.** Most people underestimate this part. You'll ask experienced teams to work differently, and doing that well is a communication skill before it's a technical one. - **A tolerance for unglamorous work.** Some of this is reconciling messy data and sitting in on a close. That's how you find out where the leverage actually is. **Nice to have** - ERP implementation or migration experience, at any level of involvement - Workflow automation platforms (Workato, Zapier, Make, n8n), ETL tooling, dbt, version control - Internal app frameworks — Retool, Streamlit, Appsmith, or you just build the thing in React and ship it - Agent frameworks, LLM APIs, retrieval, or eval tooling used in anger rather than in a side project - Experience in a SOC 2 or ISO 27001 environment, or with audit evidence and access reviews - You've been an early hire into a function that didn't exist yet ### What This Role Is Not We'd rather be direct than have you find out in month three: - **No on-call rotation and no ticket queue.** You are not corporate IT support and you're not staffing a help desk. This is a build seat. - This is not an accounting role. You won't own the close or be judged on ledger accuracy. - This is not systems administration. If the job were maintaining vendor configuration, we'd hire differently. - This is not a role where requirements arrive fully scoped. A lot of the work is figuring out what's worth building. - This is not a fast-credibility role. You're asking experienced teams to change how they work. That's earned, not assigned.