How to apply for an internal transfer to your company's AI team

Your advantage over external candidates

Your gap is usually technical depth. The plan below closes enough of it and makes the advantages visible.

Step 1: learn how the team works (weeks 1–2)

• Find out who leads the AI or data science team and what they've shipped recently. Internal newsletters, town halls and wiki pages often mention it.
• Ask for a 20-minute conversation. Say why you're interested, and ask what their hardest current problem is and what skills they look for.
• Check the internal mobility policy: minimum time in your current role, whether your manager must approve, and how internal postings are advertised.

A message that usually gets a yes:

Hi <name>, I work in <team> on <area>. I've been learning machine
learning and I'm interested in how your team works with <business area>
data. Could I have 20 minutes in the next couple of weeks to hear what
you're working on and what skills matter most? Happy to fit your calendar.

Step 2: build evidence in your current role (weeks 3–10)

Pick a problem inside your own team that uses data you're allowed to access and that the AI team would recognise as relevant.

Acceptance criteria that make it credible to a data science manager:

• A baseline to compare against — last week's value, or the current manual rule.
• A clear metric with a denominator, measured on data the model didn't see.
• A short note on limitations and what you'd do next.
• Use only data you're authorised to use, and follow your company's AI and data policies.

Keep it modest. A well-evaluated simple model is more persuasive than an ambitious unfinished one.

Step 3: show the work (weeks 10–12)

• Present the result to your own team first; their feedback strengthens it.
• Send a two-page write-up to the AI team manager you met in Step 1, and ask for their critique rather than a job.
• Offer to help with something small on their team if they need domain input.

Asking for critique does two things: you get expert feedback, and the manager starts to see you as a potential teammate.

Step 4: tell your current manager

Do this before or as you apply, according to policy. Frame it around growth:

"I've enjoyed my work here, and I'd like to grow into machine learning. I've been learning on my own time and built a small demand forecast for our products — here's what I found. The data science team may have an opening. I wanted you to hear it from me first, and I'd like to make sure my current work is well handed over if it happens."

Things to avoid: complaining about your current team, or applying without telling your manager when the policy expects it. Your current manager's support often matters to the receiving manager.

Step 5: the internal interview

Expect some of the same technical questions as external candidates — statistics, SQL, model evaluation — and prepare for them. Then lean on your advantage:

• Explain the business context of your project in terms the team's stakeholders would use.
• Mention data quirks you know about that the team might not.
• Be honest about your technical gaps and show your plan for closing them.

If it doesn't work out this time

Ask the hiring manager what would make you a stronger candidate next time, and write it down. Keep contributing where you can. Internal openings recur, and a candidate who took feedback and came back better is often remembered.

Learning alongside your job

A structured programme helps if you need statistics and machine learning foundations with feedback on your work. The Data Science programme is delivered live online, with optional in-person labs at our Noida centre, so it can run alongside a job; ask a counsellor for current batch timings. AI roles in banks, pharma and manufacturing describes how data science work differs by industry, and the skill check shows which skills your resume already evidences.