Most engineers use AI to write code. Use it to plan your move too. Copy a prompt, fill in the parts in brackets, and paste it into Claude or ChatGPT.
Turn a job description into the project that gets you hired
Paste the description of the role you actually want. Start from the job, not from a list of AI ideas.
Here is the job description for the role I want:
[PASTE JOB DESCRIPTION]
Design one production-grade project I can build alone in 4 to 8 weeks that proves every core skill in it. Include: the problem it solves for a real business, the architecture, the stack, how I would evaluate it, how I would monitor cost and failures, and three bullet points I could put on my CV once it is built.
Find the gaps between you and a role
Use it before you start learning, so you learn what the role needs and nothing else.
Here is my current experience:
[PASTE YOUR CV OR A SHORT SUMMARY]
Here is the job I want:
[PASTE JOB DESCRIPTION]
List the skills the job needs that my experience does not show yet. Rank them by how often they appear in similar AI job descriptions. For each one, tell me the smallest thing I could build to prove it, and roughly how long it would take in evenings and weekends.
Rewrite your CV for AI roles
Most engineers undersell the infrastructure work that AI teams need most.
Here is my CV:
[PASTE CV]
I am applying for [ROLE, e.g. AI Platform Engineer] roles. Rewrite my experience bullet points so they show the skills AI teams look for: running systems in production, reliability, cost, automation, data and monitoring. Keep every claim true to what I wrote. Use numbers where my text supports them, and flag any bullet that would be stronger with a number I have not given.
Practise an AI interview
Run it out loud, then paste your answers back in.
Act as a senior interviewer hiring for a [ROLE] at an AI company. Ask me one question at a time, covering production experience, system design and the tools in this job description:
[PASTE JOB DESCRIPTION]
After each answer, tell me what a strong candidate would have added, then ask the next question. After eight questions, give me my three weakest areas.
Stress-test your project like an interviewer would
Find the holes in your project before an interviewer does.
Here is a project I built:
[DESCRIBE THE PROJECT, ITS ARCHITECTURE AND STACK]
You are a staff engineer interviewing me about it. Ask me the ten hardest questions you would ask about its design decisions, failure modes, cost, scaling and security. Then tell me which of those I should be ready to answer before I put this project on my CV.
Practise an AI system design question
System design rounds are where most candidates freeze. Practise the structure.
Give me a realistic AI system design interview question for a [ROLE] role, for example a RAG platform, a model serving platform or an agent system. Let me answer step by step. Push back like a real interviewer when I skip requirements, scale, cost, evaluation, monitoring or security. At the end, show me a strong answer structure and where mine fell short.
Learn a new tool from what you already know
Learn faster by mapping new tools onto ones you have used for years.
I am a [YOUR ROLE] with experience in [TOOLS YOU KNOW, e.g. Kubernetes, Terraform, CI/CD]. Teach me [NEW TOOL, e.g. vLLM or MLflow] by comparing every concept to something I already use. Then give me a one-evening exercise that uses it in a realistic production setup.
Write your LinkedIn headline and about for AI roles
Recruiters search for the role you want, not the one you have.
Here is my experience:
[PASTE SUMMARY OR CV]
I want recruiters hiring for [TARGET ROLE] roles to find me. Write a LinkedIn headline under 220 characters and a short about section in plain, first-person language. Use the keywords those recruiters search for, keep every claim true to my experience, and avoid buzzwords.
Got a result you are not sure about?
Send me what the prompt gave you and I will tell you if it is the right project for the role you want.