About this role
Designing and building applied AI systems forms the core of the day-to-day work for this role at Coursera. Engineers in this position handle everything from initial data exploration and rapid prototyping to full solution design, experimentation, and evaluation. The team focuses on shaping how learners and enterprise customers acquire new skills on the combined Coursera and Udemy platform by turning complex datasets and ambiguous business problems into working AI solutions.
This position exists to scale AI capabilities across the newly combined organization, driving educational technology forward with modern machine learning and generative tools. Successful practitioners typically possess a deeply technical background combined with practical experience building LLM pipelines, RAG architectures, and agentic workflows. Anyone who enjoys figuring out how intelligent systems should think rather than just writing application code will find this environment well-suited to their skills.
What you will do
- Design and build complete AI systems by selecting appropriate data combinations, retrieval pipelines, agentic workflows, and structured extraction techniques.
- Architect and implement Retrieval Augmented Generation pipelines leveraging vector databases like Pinecone and Weaviate for grounded applications.
- Construct agentic AI workflows and autonomous execution loops using frameworks such as Langgraph, Mastra, LlamaIndex, CrewAI, or AutoGen.
- Design and execute comprehensive evaluation suites covering accuracy, hallucination checks, grounding, regression, and human review loops.
- Manage large-scale structured and unstructured datasets utilizing cloud-native databases and storage platforms including S3, Postgres, GCS, BigQuery, and Databricks.
- Develop rapid prototypes and experiment with diverse models, prompts, and retrieval strategies to determine optimal performance.
- Collaborate closely with product managers, data engineers, and backend or frontend engineers to translate business needs into scoped AI solutions.
- Maintain clear documentation of architectures, design choices, runbooks, and evaluation results to ensure team velocity.
What you need
- Bring at least 4 years of professional experience working in Data Science, Applied AI, or Machine Learning.
- Demonstrate hands-on experience taking AI use cases from initial data exploration through validated working prototypes.
- Possess practical knowledge of modern Generative AI techniques, large language models, embeddings, RAG, semantic search, and agentic systems.
- Complete at least two production deployments involving LLM-based systems such as fine-tuning, RAG, or agentic workflows.
- Show proven capability in designing AI experimentation frameworks, analyzing errors, and iterating based on quality metrics.
- Use Python and SQL fluently to manipulate data, build analysis workflows, and develop algorithmic solutions.
Nice to have
- Experience with Generative AI platforms and ecosystems including Vertex AI, Bedrock, Azure AI, OpenAI or Anthropic APIs, and Hugging Face.
- Background in designing vector search, tool-calling, multi-step AI workflows, and LLM gateways.
- Familiarity with AI evaluation techniques like golden datasets, LLM-as-judge, ranking metrics, and automated quality frameworks.
- Previous work with modern data platforms such as Snowflake, Spark, or Databricks.
- Knowledge of AI observability, model monitoring, responsible AI, PII handling, and prompt injection mitigation.
Skills and tech stack
Questions candidates ask
What kind of projects will I work on in this role?
You will work across data exploration, AI solution design, rapid prototyping, experimentation, and evaluation. This includes building retrieval pipelines, agentic workflows with tool use, and structured extraction from unstructured sources.
What tools and frameworks are used by the team?
The tech stack includes Python, SQL, vector databases like Pinecone and Weaviate, and agent frameworks such as Langgraph, Mastra, LlamaIndex, CrewAI, and AutoGen. Cloud-native databases and storage like S3, Postgres, GCS, BigQuery, and Databricks are also utilized.
Who will I be collaborating with on a regular basis?
You will work closely with Product Managers, Data Analysts, and Software Engineers. Additionally, you will partner directly with the business teams who use the AI applications you build.
How can I ensure my application is safe from recruitment fraud?
Recruiters from Coursera and Udemy only communicate via official coursera.org or udemy.com email addresses and never through personal accounts. Applications must be submitted directly through the official careers page.