# Writing a CV for AI Roles | LLMnet Jobs Share:[𝕏](https://twitter.com/intent/tweet?url=https%3A//vacatures.llmnet.nl/cv-schrijven-voor-ai-functies&text=CV%20schrijven%20voor%20AI-functies)[LinkedIn](https://www.linkedin.com/sharing/share-offsite/?url=https%3A//vacatures.llmnet.nl/cv-schrijven-voor-ai-functies)[Reddit](https://www.reddit.com/submit?url=https%3A//vacatures.llmnet.nl/cv-schrijven-voor-ai-functies&title=CV%20schrijven%20voor%20AI-functies)[Facebook](https://www.facebook.com/sharer/sharer.php?u=https%3A//vacatures.llmnet.nl/cv-schrijven-voor-ai-functies)[Copy link](#) # Writing a CV for AI Roles By Ivo Donker — compiled with AI support (Claude & Gemini) Writing a CV for roles in artificial intelligence (AI) and Large Language Models (LLMs) requires a specific approach. Where regular IT roles often rely on well-known frameworks and standard programming languages, evaluation for AI-related roles is about demonstrating control, manageability, and practical effectiveness of the systems you build or implement. A strong AI CV bridges the gap between theoretical knowledge and operational reality. In this article, we cover the different readers of your CV, how to substantiate your projects, presenting lateral entry honestly, and ways to effectively integrate a portfolio. ## The Two Reading Rounds: Corporate Recruiter versus Technical Reviewer When applying for an AI role, your CV consistently passes through two substantively different filters. To get through the selection process, the document must provide the right information for both profiles. ### Round 1: The Recruiter and ATS In the first phase, a recruiter (or an Applicant Tracking System) reviews your document. This reader primarily looks at the big picture, minimum education requirements, relevant job titles, and key terms from the job posting. The recruiter wants answers to questions such as: Does this candidate have experience with Python? Have they worked with cloud infrastructure? Does the profile match the expected seniority level? If you want to align your salary expectations with the market standards that recruiters use, you can consult the market comparisons on our page about [salaries for AI roles in the Netherlands](/en/salarissen-ai-rollen-nl). ### Round 2: The Tech Lead or Engineering Manager Once the CV passes the initial screening, it reaches the domain specialist: a Lead AI Engineer, Data Science Manager, or CTO. This reader doesn't scan for isolated keywords but looks for technical depth, decision-making, and practical constraints. The technical reviewer asks themselves questions such as: Does this candidate understand the difference between calling an API and building a robust evaluation pipeline? How did they handle latency, costs, and data security? Key insight: Your CV should be structured so that the recruiter recognizes the required skills within ten seconds, while the technical reviewer immediately reads the architectural choices and constraints in the work experience descriptions. ## From Tool Overview to Concrete Project Context Many applicants make the mistake of merely listing a collection of technologies under experience (such as "Python, LangChain, OpenAI, Vector DBs"). A list doesn't prove professional competence. A technical reviewer wants to understand what your specific contribution to the solution was. To describe a project or work experience compellingly, use a fixed format consisting of five elements: - The problem: What business or technical challenge needed to be solved? - The approach: Which architecture, models, or methods were chosen and why? - Your role: What exactly was your individual contribution (versus the team's)? - Measurable result: What impact did the solution have on performance, costs, or processes? - Reflection: What would you have designed differently technically with the knowledge you have now? By applying this framework, a passive sentence like "Worked on a RAG system" transforms into a clear, substantive description of responsibility. ## Demonstrable Responsibility in Production There is a fundamental difference between building a prototype in a Jupyter Notebook and running an AI application in a production environment. Your CV must demonstrate that you have an eye for the risks and challenges that arise once real users interact with a model. Make sure your work experience clearly shows responsibility in the following four domains: ### 1. Evaluation and Validation Describe how you measured the quality of the output. Did you use automated evaluation frameworks, human-in-the-loop review loops, or specific benchmarks? Indicate how you kept hallucinations or incorrect output within acceptable limits. ### 2. Cost Management (FinOps) LLM applications can become costly at high request volumes. Mention how you optimized token usage, for example by implementing caching, choosing smaller or specialized models, or clever prompt architecture. ### 3. Reliability and Latency Production systems require low response times and high availability. Mention measures such as asynchronous task processing, streaming responses, or setting up fallback mechanisms when an API service goes down. ### 4. Security and Privacy How were sensitive business or personal data handled? Show that you are familiar with data anonymization, preventing prompt injections, and complying with privacy legislation. ## Presenting Lateral Entry and Reskilling Honestly Many professionals make the transition to the AI domain from related disciplines, such as software engineering, data analysis, system administration, or even non-technical domains. Transparency is essential here. Don't pretend you have years of experience as an AI Engineer if you only recently made the switch. Instead, focus on your transferable skills. An experienced software developer who has reskilled brings valuable knowledge of CI/CD, testing, code quality, and software architecture — aspects that are often still missing in junior AI specialists. If you are following a reskilling trajectory, also read our advice on [reskilling into AI](/en/omscholen-naar-ai) to make the most of your accumulated domain knowledge in your positioning. ## Integrating a Portfolio and Repositories A link to a GitHub profile or online portfolio at the bottom of your CV is often overlooked by recruiters. If you want a technical reviewer to actually look at your code or projects, you need to keep the barrier as low as possible. Include a specific, contextual reference directly with the relevant work experience or project description. Don't simply write "GitHub," but indicate what the reader will find there (for example: "See /repository-name for the evaluation scripts and latency benchmark"). For a detailed explanation of the structure of a representative projects page, you can consult the guide on [building an AI portfolio](/en/ai-portfolio-bouwen). If you want to study in-depth articles about LLM architectures for your own projects, check out the tutorials on [leren.llmnet.nl](https://leren.llmnet.nl/en/). ## Common Mistakes on an AI CV During the selection process, many candidates are eliminated due to avoidable content and formatting mistakes. The four most common pitfalls are: - Lists of scattered buzzwords: Listing dozens of AI terms without context raises doubts about the true depth of your knowledge. - Certificates as the main course: Course certificates demonstrate a willingness to learn, but they don't replace practical experience. Place certificates in a separate, compact section at the bottom of the CV and prioritize building and implementation experience. - Inflated role descriptions: Claiming that you trained an entire model while merely calling an existing API with a basic prompt is immediately exposed during a technical interview. In preparation, check the structure of a [technical assessment](/en/technisch-assessment-voorbereiden) to see how this knowledge is tested. - Unreviewed AI-generated CV texts: Using generative AI to write your CV is allowed, but unreviewed output is immediately recognizable by excessive jargon, vague catch-all terms, and a lack of concrete facts. Always review and rewrite AI-generated texts yourself. ## Strong versus Weak Phrasing The table below shows the difference between vague, passive statements and concrete, substantive phrasing on a CV. Weak (non-committal / buzzwords) | Strong (concrete / substantive) | Gained experience with LLM integrations and vector databases. | Set up a RAG pipeline with Qdrant and LlamaIndex to make 50,000 internal policy documents searchable. | Responsible for optimizing AI prompts. | Revised the prompt structure and added few-shot examples, reducing the percentage of incorrect JSON output by 40%. | Conducted research into the use of AI within the department. | Benchmarked three open-source models on processing speed and accuracy to replace a costly external API service. | Developed a chatbot using OpenAI. | Built a customer service assistant via the OpenAI API, including streaming responses and a Python-based fallback module for time-outs. | ## Tailoring Your CV to Each Job Posting Without Wasting Time It's unwise to write a completely new CV for every application. However, it's equally unwise to send exactly the same document to every employer. The most efficient approach is to maintain a 'Master CV'. In this Master CV, you document all projects, responsibilities, technologies used, and measurable results in detail. When you respond to a specific job posting, make a copy and only adjust the emphasis: - Rewrite the summary (profile text) at the top so it aligns with the specific role (for example, more focus on MLOps or more focus on application development). - Reorder the bullet points under your most recent work experience: put the points that directly match the job requirements at the top. - Select the two or three most relevant projects or repositories and highlight them explicitly. ## Checklist Before Submitting Review your CV against this checklist before sending your application: - Is the CV easily readable for both a recruiter (quickly scannable) and a technical reviewer (sufficient depth)? - Are projects described with clear mention of the problem statement, your own role, and the result achieved? - Does the description of AI systems include information about evaluation, cost management, or reliability in production? - Are any career changes or reskilling trajectories presented honestly with an emphasis on transferable skills? - Do all hyperlinks to GitHub, a portfolio, or specific repositories work correctly? - Is the text free of vague buzzwords and excessive AI-generated hollow phrases? - Is the profile text and the order of experiences aligned with the specific job posting?