Company description
Spark Foundry was built during the most transformative time in the history of advertising and marketing. We deliver everything a traditional media agency has to offer and have grown into one of the world’s most successful agencies by helping brands evolve their marketing by leveraging identity, commerce, artificial intelligence, and more to connect with people. Transformation is part of our DNA. Spark Foundry’s start-up spirit delivers high-touch approaches and a never-ending desire to challenge the status quo. Combined with Publicis Media’s powerhouse soul leveraging connected data assets, advanced AI applications, and investment clout, we “Bring HEAT to Brands.” No other agency possesses the expertise that we do to address today’s most pressing challenges to drive business transformation through media.
Overview
As an ML/AI engineer, you'll develop and deploy both GenAI applications and traditional ML systems that help media planners, strategists, and analysts work smarter. You'll build models that forecast campaign performance, LLM-powered tools that generate insights, and recommendation systems that optimize media strategies. This is a hands-on role where you'll own projects end-to-end and see your work used daily by agency teams.
Responsibilities
- Design, build, and deploy machine learning and GenAI solutions for media use cases
- Develop time-series forecasting, classification, clustering, and recommendation models, taking them from experimentation through production deployment
- Build LLM-powered applications such as RAG systems, conversational analytics, insight generators, and prompt-based tools for content and creative workflows
- Integrate and operationalize foundation models (e.g., OpenAI, Anthropic Claude, etc.), including prompt design, vector search, semantic retrieval, and output optimization
- Own ML features end-to-end from requirements gathering and experimentation to production deployment, monitoring, and iteration
- Write production-quality, scalable code with proper testing, documentation, performance optimization, and error handling
- Monitor deployed models for accuracy, drift, reliability, and business impact; retrain and improve models based on real-world usage
- Experiment rapidly, evaluate tradeoffs across accuracy, cost, latency, and complexity, and validate solutions with real users
- Collaborate closely with data engineers and agency teams to align data pipelines, technical solutions, and business needs
- Communicate technical concepts clearly to non-technical stakeholders and contribute to documentation, reviews, and knowledge sharing
- Stay current on ML and GenAI advancements relevant to media, evaluate new tools pragmatically, and bring informed recommendations to the team
Qualifications
- 2–4+ years of hands-on experience building and deploying ML models in production
- Strong Python skills and experience with ML frameworks (scikit-learn, PyTorch, TensorFlow, XGBoost, LightGBM)
- Practical experience building applications with LLMs (OpenAI, Anthropic, or open-source models)
- Solid understanding of ML fundamentals: feature engineering, evaluation, overfitting, imbalanced data
- Experience deploying models beyond notebooks (APIs, batch jobs, real-time inference)
- Strong SQL skills and experience working with large datasets
- Experience with cloud platforms (AWS, GCP, or Azure) and Docker
- Software engineering fundamentals: Git, testing, CI/CD, code reviews
- Ability to work independently, make pragmatic tradeoffs, and navigate ambiguity
- Experience with GenAI architectures: RAG, embeddings, vector databases (Pinecone, Weaviate, Chroma)
- Familiarity with LangChain, LlamaIndex, or similar orchestration frameworks
- Experience with time-series forecasting, recommendation systems, or ranking models
- Exposure to MLOps practices (MLflow, monitoring, feature stores)
- Experience with distributed data processing (Spark, Databricks)
- Understanding of experimentation and A/B testing
- Familiarity with media and advertising concepts (reach, frequency, attribution)
Character:
The following qualities help drive success as member of the Spark Data and Analytics team:
- Entrepreneurial, engaging, resourceful, curious, and self-directed spirit
- Willing and easily roll sleeves up or down; love the nitty-gritty and the strategy
- Collaborative approach to building cohesive, strong teams
- Loving and living the intersections between brands, people, media, communications
- Relentlessly passionate and resolute
- Planning and time management excellence.
- Embrace challenges
- Proactive, especially in pushing for new opportunities, approaches, and ideas.
- Keenly focused on action and solutions; thrives with deep critical thinking and analysis.
- Pioneering insight attitude and research in-the-know.
- Resourcefulness, flexibility and adaptability, strong ability to pivot when the need arises.
- Inspired to be part of the insight journey/revolution with a growing