AI Operations Analyst
2 days ago
Birmingham
About the role: Apt Companies is an IT staffing and workforce solutions firm. Over the past year we've built a real internal tech stack: a custom platform called Apt Command, integrations into our ATS and our data providers, and a growing pile of automations across operations, sales, and marketing. One person built all of it. He isn't a software engineer, and we're not hiring one now. We have more ideas in flight than hands to finish them. Projects get built to eighty percent and then parked when something urgent shows up. Reports still get assembled by hand. Recruiters still retype data that already lives in another system. And the tooling in this space changes every few weeks, so keeping current with it is close to a job by itself. You'd own that backlog. Take projects from idea to shipped to actually in use, and keep hunting for the automations that bring in revenue rather than the ones that only save time. You'll report to our Director of Operations, learn the staffing business from the inside, and use AI coding tools the way an analyst uses Excel. What you'll do: First 90 days • Sit with recruiting, delivery, sales, and back office. Learn how staffing works end to end, req through submittal through placement through invoice, and find where the manual steps actually are rather than where we assume they are, • Get fluent in our stack: Apt Command, our ATS and its integrations, our data providers, our automation tooling, • Pick two or three recurring reports or manual workflows and rebuild them so they run themselves, • Ship something people use in your first month, even if it's small Months 3 to 9 • Own the internal AI and automation roadmap: intake requests from the business, size them, sequence them, drive them to done, • Finish the projects sitting at eighty percent. A good chunk of this job is exactly that, and we'd rather say so now than have you find out in month two, • Build and extend internal tooling: reporting, dashboards, data hygiene, workflow automation, • Build agent workflows that support outbound work: research, enrichment, list building, content production, sequencing, • Train the team on what you build and check whether they're using it. Adoption is how we'll judge the role, not how much got shipped Months 9 and on • Keep up with the market: AI tooling, agent frameworks, what other staffing firms are and are not doing. Come back with proposals rather than links, • Find automation that generates revenue: earlier signal on hiring demand, better candidate matching, higher submittal quality, more outreach at the same headcount, • Build validated write-back paths so data entered once lands correctly in the system of record, • Take on more scope as you earn it. This role should grow into leading internal AI and automation for the firm What we're looking for: Required • Some real experience owning a messy process. Operations, recruiting or staffing ops, sales ops, analytics, marketing ops, project management, back office. Anywhere things broke and you were the one who fixed them. We're not counting years, • Advanced with spreadsheets, well past VLOOKUP. You've built things in Excel or Sheets that other people depend on, • Something you built. A working tool, a live automation, a bot, a side project, a process you rebuilt. Built at work or built at home, we don't care which. We'll ask you to walk us through it in detail, including the parts that broke, • Hands-on time in an AI coding environment like Claude Code, Cursor, or Codex. Not just chatting with a model, actually building with one, • Data sense. When a report double-counts, you can work out why. You don't need to know SQL yet, • Judgment about what to build. You can tell what someone asked for apart from what they need, and you'd rather ship a rough tool people use tomorrow than a polished one in six months, • A habit of finishing. Error handling, documentation, training, and follow-up are the last twenty percent, and they are the job, • Self-direction. Nobody here is going to hand you a ticket queue, • Enough humility to be told by a recruiter that you solved the wrong problem, and go rebuild it Helpful, not required • SQL, • Staffing, recruiting, or professional services background, • Bullhorn, Salesforce, or another ATS or CRM, • Workflow automation tools like n8n, Zapier, or Make, • BI and dashboarding experience, • APIs, webhooks, JavaScript, Python, or any coding at all, however informal Not required • A computer science degree, • Professional software development experience, • A specific number of years in anything, • Having written code before this year How the process works: Resumes are a weak filter for this role, so here is what to expect... • Apply with proof. Skip the cover letter. Send us a link, a screenshot, a short screen recording, or two paragraphs about something you built and why. This is the real first screen, and it matters more to us than your resume, • Intro call. Your background and what you've built, • Show us something. Screen share a tool, automation, or process you built with AI. We want to see how you worked through it, where it broke, and how you figured out the fix, • Working session, onsite. We'll give you a real Apt problem and a few hours. Use whatever tools you normally use. We're watching how you break down something unfamiliar, not whether you finish, • Conversation with leadership. Judgment, communication, and why you want the job Guardrails We want you moving fast, and AI-generated code still needs verification. So a few rules: • Read-only integrations before write access, • No production credentials in code, and no shared logins, • Candidate and client data is confidential and in places regulated. It doesn't go into unapproved tools and it doesn't leave our systems, • Anything touching payroll, invoicing, or client-facing reporting gets reviewed before it goes live, • Sandbox and test environments before production, • Document what you build in plain language, as you build it. Someone else should be able to run it without you in the room Who this suits Someone who has been quietly automating their own job with AI and wants to do it full-time, with real backing and without having to ask permission. Someone who reads about this on their own time and likes finishing things. We want you to grow here. This role should turn into leading internal AI and automation for the firm, and we'd rather build that person than go hire them. Probably not a fit if you want a defined ticket queue, if you'd rather write strategy decks than build, if you'd rather not talk to the people who will use what you make, or if you need a senior engineer to check your work before you feel comfortable moving.