Yes, this transition is very realistic and in high demand right now. With 12 years of software engineering experience (especially mobile apps), you already have a massive advantage that most junior AI candidates lack: strong software architecture, production deployment knowledge, debugging complex systems, and an understanding of user-facing applications. London’s AI job market in 2026 is booming—particularly for AI/ML Engineers, Generative AI Engineers, and MLOps-focused roles in fintech, tech consultancies, and scale-ups. Roles often pay £80k–£130k+ and heavily value people who can ship reliable, production-grade AI systems (not just train models).

The gap from mobile (Swift/Kotlin/Java) to AI is bridgeable in 6–12 months of focused part-time effort (or 3–6 months full-time). The key is to treat this as an engineering upgrade, not a complete career restart.

1. Quick Skills Gap Analysis (What London Employers Want in 2026)

From current job postings (Indeed, Glassdoor, LinkedIn, Built In London):

Biggest 2026 trend: Generative AI + production deployment (not just research). Employers want engineers who can turn an LLM into a reliable product.

2. 6–12 Month Preparation Roadmap (Tailored for Experienced Devs)

Month 0 (1 week): Audit your skills

• Do a quick Python refresher + basic ML (free: Andrew Ng’s “Machine Learning” on Coursera – still the gold standard).
• Set up a GitHub portfolio repo now.

Months 1–2: Foundations (10–15 hrs/week)

• Python for Data Science + ML (NumPy, Pandas, scikit-learn).
• Math refresh: 3Blue1Brown linear algebra + StatQuest probability (YouTube).
• Course: Andrew Ng “Machine Learning Specialization” or “Deep Learning Specialization” (Coursera).
• Milestone: Build 2–3 classic ML projects (e.g., classification, regression).

Months 3–5: Core AI Engineering + GenAI (the money skills)

• PyTorch (preferred) + Hugging Face.
• Deep Learning: Fast.ai (practical) or Deep Learning Specialization.
• GenAI stack (this is what gets you interviews): Prompt engineering → RAG → Agents → Evaluation.
  • Use free resources: LangChain/LlamaIndex tutorials, “Building LLM Apps” courses on DeepLearning.AI or DataCamp.
• Cloud: Pick one (AWS or Azure – both have strong London presence) and get the Practitioner/Associate cert.

Months 6–8: MLOps + Production

• Docker + Kubernetes basics.
• MLflow, CI/CD for models, monitoring (Prometheus/Grafana).
• Deploy 2–3 models end-to-end (e.g., to AWS SageMaker or Azure ML).

Months 9–12: Portfolio + Job Hunt

• Build 4–5 polished projects (see below).
• Contribute to 1 open-source AI repo or internal AI feature at work.
• Start applying (target 10–15 applications/week).

3. Portfolio Projects That Will Make You Stand Out (Leverage Your Mobile Background)

Recruiters want production thinking, not toy notebooks. Build these and deploy them (Streamlit/FastAPI + Vercel/Hugging Face Spaces or cloud):

1. AI-Powered Mobile Feature (your superpower) – On-device image classification or object detection app (TensorFlow Lite / Core ML / MediaPipe). Show how you optimised for mobile constraints.
2. RAG Chatbot (“Ask My Docs” or company knowledge base) – Most in-demand project in 2026. Use vector DB (Pinecone/Chroma), hybrid search, and evaluation metrics.
3. Multi-Agent Workflow – E.g., an AI research agent that uses tools (search, calculator, code interpreter) to complete a task end-to-end.
4. LLM-Powered Mobile Backend Service – Backend API that powers smart features in a mobile app (recommendations, smart replies, summarisation). Deploy with FastAPI + cloud.
5. Production ML Pipeline – End-to-end: data → training → evaluation → deployment + monitoring (include A/B testing or drift detection).

Host everything on GitHub with excellent READMEs (architecture diagrams, metrics, lessons learned). This is what gets you past ATS and recruiter screens.

4. Certifications (HR Filters + Credibility)

• Strongest for London: Google Professional Machine Learning Engineer or AWS Certified Machine Learning – Specialty.
• Faster options: AWS AI Practitioner or Azure AI Engineer Associate (AI-102).
• Free/cheap signal: Andrew Ng courses + Google Prompting Essentials.

5. Resume & LinkedIn Optimisation

• Lead with “Senior Software Engineer transitioning to AI Engineering” (or “AI Engineer with 12+ years production software experience”).
• Quantify everything: “Built and shipped mobile apps used by 500k+ users” → “Designed AI-enhanced mobile features reducing latency by 40% using on-device inference.”
• Add a “AI Projects” section prominently.

• LinkedIn banner: “Mobile → AI Engineer

  Building production GenAI systems in London”.

• Turn on #OpenToWork (green banner).

6. Networking in London (Critical – Referrals Win Jobs)

London’s AI scene is very accessible:

• Meetups: AI Professionals Central London, London AI Developers Group (Meetup.com).
• Major events 2026: AI Engineer Europe (April 8–10), Generative AI Summit, MLcon London, AI & Big Data Expo.
• Smaller networking: Eventbrite “AI & Tech Networking London” events (Old Street/Shoreditch area).
• Reach out to recruiters at Harnham (specialist Data/AI London recruiter) and on LinkedIn.
• Join London AI Slack/Discord communities.

7. Job Search Strategy

• Platforms: LinkedIn (set location to London + remote UK), Indeed, Glassdoor, AIJobs.ai, Built In London.
• Target companies: DeepMind, Google, Meta, Anthropic (London offices), fintech (Revolut, Monzo, etc.), consultancies (Deloitte, Accenture AI practices), scale-ups.
• Apply to AI Engineer / ML Engineer / Generative AI Engineer titles. Many accept strong software engineers with a good portfolio.
• Recruiters move fast—have your portfolio link ready.

Final Tips & Realistic Timeline

• Daily habit: 1–2 hours coding + 30 min reading (arXiv sanity or Hugging Face blog).
• Side project rule: Ship something every 2–3 weeks.
• Interview prep: LeetCode (medium/hard Python), ML system design (Grokking the Machine Learning Interview or “System Design for Interviews” resources), and behavioural stories from your mobile career.
• Salary expectation: With your experience + solid portfolio → £90k–£120k+ realistic in first AI role.

You’re not starting from zero—you’re upgrading a very strong foundation. Many people with your exact background have made this switch successfully in under a year. Start today with Python + the first Andrew Ng course, and in 3 months you’ll already feel the momentum.

If you want a personalised 30-day starter plan, specific course links, or help reviewing your resume/projects, just share more details and I’ll refine this further. You’ve got this! 🚀