My Six-Month Career Transition Plan from Office Administration to AI Operations — What Am I Missing?

I’ve been an office administrator for seven years. I’m good at it. I know how to manage calendars, coordinate vendors, keep records straight, onboard new hires, and hold together the thousand small processes that make an office actually function.

But I’ve also watched AI enter my workflow over the past 18 months. It drafts emails. It summarizes meetings. It organizes spreadsheets. It answers policy questions. It schedules. None of it is perfect, but it’s good enough that I can see the writing on the wall: the administrative roles that survive will be the ones that manage AI, not the ones that compete with it.

So I’m building a six-month transition plan from office administration into AI operations.

Not AI engineering. Not data science. Not prompt engineering as a job title. I mean the operational layer: the person who helps a team adopt AI tools, map workflows, write prompts and SOPs, train colleagues, track quality, manage vendors, and make sure the tools actually solve real problems instead of creating new ones.

I’m sharing my plan here because I want feedback. What am I missing? What’s unrealistic? What would you add, cut, or reorder? If you’ve made a similar transition — from admin, ops, customer support, or another non-technical role into AI-adjacent work — I want to hear what actually worked.

What I Mean by “AI Operations”

One problem I’ve run into: “AI operations” is not a standardized job title. It shows up under different names at different companies. Here are the role titles I’m targeting:

• AI Operations Specialist

• AI Enablement Specialist

• AI Adoption Manager

• Automation Operations Analyst

• AI Workflow Coordinator

• AI Program Analyst

• AI Quality & Training Specialist

• AI Operations Coordinator

The common thread is that these roles sit between the AI tools and the people using them. They’re not building models. They’re building the systems, habits, and guardrails that make AI useful at work.

Typical responsibilities I’ve seen in job postings:

• Evaluating and piloting AI tools for specific teams

• Mapping workflows before and after AI adoption

• Writing prompt libraries, SOPs, and usage guidelines

• Training employees on AI tools and responsible use

• Tracking quality, errors, and time savings

• Coordinating with vendors, IT, legal, and HR

• Managing AI-related change and adoption

• Maintaining a human-in-the-loop review process

That’s a role I can grow into from administration. It uses skills I already have — coordination, documentation, process improvement, training — and adds AI fluency on top.

My Six-Month Plan

Here’s the plan I’ve built. It’s not a full-time bootcamp. I’m doing this around a full-time job, so I’ve tried to keep it realistic: 5–7 hours a week, plus whatever I can learn on the job.

Month 1: Baseline and Target Roles

Goal: Understand where I am and where I’m going.

Actions:

• Audit my current job: list every recurring task, how long it takes, and which parts AI already touches.

• Collect 20 job postings for AI operations, AI enablement, automation operations, and AI adoption roles.

• Highlight the skills that appear most often: project management, change management, prompt writing, data analysis, training, governance.

• Pick three target role titles and three target industries (I’m leaning toward healthcare, education, and professional services).

• Set up a simple portfolio document where I’ll store everything I build.

What I’m reading: WEF Future of Jobs Report 2025, SHRM’s AI in the workplace research, and job postings themselves. Job postings are the most honest curriculum.

What I’m missing: I don’t know if I should target a specific industry or stay general. Advice welcome.

Month 2: AI Tool Fluency

Goal: Get hands-on with the tools I’ll be expected to manage.

Actions:

• Use ChatGPT, Claude, Copilot, and Gemini for real work tasks — not just experimentation.

• Learn one automation platform: Zapier, Make, or Power Automate. I’m starting with Zapier because it’s common in ops roles.

• Build a prompt library for administrative workflows: meeting summaries, vendor comparisons, policy drafts, onboarding checklists, spreadsheet cleanup.

• Keep a correction log: every time AI gets something wrong, note the task, the error, and how long it took to fix.

• Document five workflows with before-and-after time estimates.

Deliverable: A prompt library and a correction log I can show in interviews.

What I’m missing: How deep do I need to go on automation tools? Is Zapier enough, or do I need Make and Power Automate too?

Month 3: Operations and Process Skills

Goal: Build the operational backbone that makes AI useful.

Actions:

• Learn process mapping. I’m using Lucidchart and Miro to map three office workflows.

• Write two SOPs: one for AI-assisted meeting notes, one for AI-assisted vendor research.

• Take a project management course. I’m considering Google Project Management or CAPM. I already do project coordination, but I want the vocabulary and credential.

• Learn Notion, Asana, or Monday at a deeper level. These are the systems where AI ops work often lives.

• Build one small automation demo: e.g., a form submission that triggers an AI summary and adds it to a spreadsheet.

Deliverable: A workflow map, two SOPs, and one automation demo.

What I’m missing: Do I need SQL or data visualization? Some job postings mention it, but I don’t know if it’s required or preferred.

Month 4: Governance and Risk

Goal: Understand the guardrails that make AI safe and trustworthy at work.

Actions:

• Study the NIST AI Risk Management Framework. It’s the closest thing the U.S. has to a shared language for AI risk.

• Learn the basics of data privacy: what can and cannot be entered into AI tools, how to handle PII, and what “human-in-the-loop” means in practice.

• Draft a model AI usage policy for a hypothetical team. Include: approved tools, data rules, disclosure expectations, review process, and escalation path.

• Read about AI bias, hallucination, and over-reliance. SHRM’s 2026 research found that 80% of HR professionals believe repeated AI drafting reduces original thinking — that’s a governance issue, not just a training issue.

• Interview one person in IT, legal, or compliance about how AI policies are actually enforced.

Deliverable: A draft AI usage policy and a one-page summary of key risks.

What I’m missing: How much legal knowledge do I actually need? I don’t want to pretend to be a lawyer, but I also don’t want to be blindsided by compliance questions.

Month 5: Portfolio and Proof

Goal: Turn my learning into evidence I can show.

Actions:

• Write three case studies:

  1. How I automated meeting notes and saved X hours per week.

  2. How I used AI to speed up vendor comparison and what errors I caught.

  3. How I mapped a workflow before and after AI adoption.

• Quantify everything: time saved, error rates, number of steps reduced, people trained.

• Publish the case studies on LinkedIn or a simple personal site. Keep them concise and outcome-focused.

• Volunteer to help with an AI pilot at my current job, even if it’s small. Real experience beats hypothetical projects.

• Ask two colleagues to review my portfolio and tell me what’s confusing.

Deliverable: Three case studies with metrics and a portfolio page.

What I’m missing: How do I write case studies without violating confidentiality? I’ll anonymize company details, but I’m not sure where the line is.

Month 6: Job Search and Networking

Goal: Convert the plan into interviews.

Actions:

• Rewrite my resume around outcomes and AI-adjacent skills, not administrative tasks.

• Create a target list of 30 companies: AI-forward, mid-sized, with operations or enablement roles.

• Do 10 informational interviews. Ask: “How did you get into this role?” “What skills matter most?” “What would you want someone in this role to know on day one?”

• Apply to 20 roles. Track responses and refine my materials based on what gets interviews.

• Practice interview questions: “How would you evaluate an AI tool for a team?” “How would you handle an employee who refuses to use AI?” “How do you measure whether AI is actually helping?”

• Join communities like The Workshift Forum, relevant Slack groups, and local ops meetups.

Deliverable: 20 applications, 10 conversations, and at least 3 first-round interviews.

What I’m missing: I don’t know how to explain the transition without sounding like I’m running away from administration. I want to frame it as a deliberate move toward AI operations, not away from admin.

What I’m Doing Alongside the Plan

Documenting everything. I keep a running log of tasks, tools, errors, and time savings. This is my evidence base.

Talking to people. I’ve already had two conversations with people in AI enablement and automation operations. Both said the same thing: the role is 70% operations, 30% AI. That was encouraging.

Staying employed. I’m not quitting my job. I’m using it as a lab. Every process I improve, every tool I test, every SOP I write is portfolio material.

Being honest about the timeline. Six months may be optimistic. I’m treating it as a structured start, not a guarantee. If it takes nine months, that’s fine — as long as I’m moving.

What I Think I’m Missing (Please Tell Me)

Here’s where I need help. These are the gaps I can see, and I suspect there are more I can’t.

1. Technical depth. Do I need SQL? Python? API basics? Or can I succeed in AI operations with no-code tools and strong process skills? I don’t want to spend months learning to code if it’s not necessary, but I also don’t want to hit a ceiling.

2. Certifications. Which ones actually matter? I’ve seen Google Project Management, CAPM, HubSpot, and various AI certificates. Do any of them move the needle, or are they just resume filler?

3. Portfolio without a title. How do I prove I can do AI operations when I’ve never had the title? Are case studies enough? Do I need a personal website? Should I publish on LinkedIn?

4. Industry focus. Should I go deep in one industry — healthcare, education, finance — or stay general? Does specialization help or limit me?

5. Salary expectations. I’m willing to take a small step back to change direction, but I don’t want to undersell myself. What’s a realistic range for an entry-level AI operations role?

6. Remote vs. on-site. Many AI operations roles seem to be hybrid or on-site. Is remote realistic for someone breaking in, or should I expect to be in an office?

7. Internal transfer vs. external move. Should I try to create an AI operations role at my current company, or apply externally? Internal might be easier, but it might also be slower.

8. The “admin” label. How do I avoid being pigeonholed as an administrative assistant when I apply for AI operations roles? What language should I use?

9. Keeping up. AI changes every week. How do I stay current without drowning in hype? What sources do you actually trust?

10. What I haven’t considered. This is the big one. What am I not asking? What blind spots do you see?