ML Engineer

Marketing.MBA: Digital Marketing Powerhouse
Summary
Join Marketing.MBA, a marketing agency building market leaders, as their Machine Learning Engineer. You will build the core ML infrastructure for their proprietary marketing AI systems, focusing on developing and fine-tuning custom LLMs for encoding marketing strategies. This foundational role in their AI Lab involves designing pipelines, creating training data, and implementing fine-tuning approaches. You will collaborate with the Head of AI and strategy team, tackling challenges in teaching AI about persuasion and brand psychology. The position offers a fast-paced, creative environment with opportunities for professional development and flexible work arrangements. Competitive salary, bonuses, and additional perks are included. Become part of a team revolutionizing AI-driven marketing strategy.
Requirements
- Proven, hands-on experience fine-tuning LLMs (e.g. Llama 2/3, Mixtral) with Hugging Face’s Transformers, TRL, PEFT, and related tools
- Strong Python development background with emphasis on data engineering, model evaluation, and pipeline design
- Deep understanding of fine-tuning techniques like QLoRA, DPO, and SFT, and the tradeoffs they entail
- Mid-level to senior-level software engineering experience, including Git, testing, Docker, and reproducible ML practices
- Ability to independently debug complex training and inference issues and drive improvements
- Solid grasp of evaluation methods for generative AI, prompt design, and model alignment
- Interest in applying AI to nuanced and human-centric challenges like branding, psychology, and creativity
- Strong communication skills and ability to translate between strategic intent and technical implementation
- Fluent in English
- Ability to join the team within 2 weeks of offer to support product timelines
Responsibilities
- Own and build the entire ML pipeline — from raw data extraction and preprocessing to model training, fine-tuning, evaluation, and deployment
- Develop and maintain a high-throughput data factory, distilling strategic insights from deep-dive marketing analyses into synthetic training data
- Train and fine-tune open-source LLMs (Llama 3, Mixtral, etc.) using modern techniques like SFT, DPO, QLoRA, and Hugging Face’s TRL/PEFT libraries
- Partner closely with the Head of AI and strategy team to transform abstract marketing frameworks into formal, teachable representations for models
- Engineer robust, scalable systems for model reproducibility, experiment tracking, and deployment efficiency
- Tackle novel challenges in teaching AI concepts such as brand positioning, tone, and persuasive messaging
- Rapidly iterate across experimentation cycles to evaluate performance and unlock model improvements
- Contribute to our proprietary model training workflows and support integration with internal agent systems
- Collaborate in weekly syncs (1x/week) to stay aligned on vision, progress, and shared technical goals
- Lay the groundwork for our own family of specialized marketing LLMs—optimized for reasoning, latency, and quality
Preferred Qualifications
- A portfolio or GitHub showcasing LLM fine-tuning or NLP experimentation projects
- Familiarity with model distillation or synthetic data generation
- Experience training models for creative, strategic, or high-stakes decision-making domains
- Background in marketing, behavioral science, or persuasive technologies
Benefits
- Remote & Flexible : Enjoy the flexibility and trust that come with a remote work environment and flexible working hours, ideally in CET timezone +/- 2hrs
- Attractive Compensation : Benefit from competitive salary packages, bonuses, and additional perks that recognize and reward your performance
- Lifelong Learning : We invest in your professional development. Benefit from our extensive online course library to continuously grow your skills