Junior Developer, Data & AI

Turn raw transactions into answers retailers act on, and build the AI that explains them

Your code goes to production, not to a backlog. At AppCard, the data you model powers the reports retailers use to run their business. The assistant you help build answers real merchants’ questions in plain English. You’ll see your work used within weeks.

Why engineers like working here

• A modern data stack, not a legacy one. You’ll work with a semantic layer, dbt-style transformations, an orchestrator, and BI tooling. These are the skills the market is hiring for.
• Real AI work, not a demo. You’ll help build a natural-language analytics assistant over live business data. You’ll tune prompts, design evaluation sets, and learn why LLM systems fail and how to measure and fix that. Most juniors don’t get this exposure.
• Real data at real scale. millions of shoppers / thousands retailers / terrabytes of data. Correctness matters here, and you’ll learn to find the bug hiding in the numbers.
• Ownership early. In 90 days you’ll own a pipeline or model end to end, from design through monitoring.
• A named mentor and a 30/60/90 plan. You won’t be left to figure it out alone.
• Small team, big surface area. You’ll work next to senior engineers, and your opinion counts in design discussions.
• Code review that teaches. Every change is reviewed by someone who’ll explain the “why”, and you’ll review others too.
• We encourage using AI tools well. We expect you to understand what ships, and we’ll teach you how.

What you’ll build

• Data models and pipelines that turn raw transactions into trustworthy metrics
• Semantic-layer definitions that power dashboards and the assistant alike
• Evaluation harnesses that tell us whether the AI got better or just sounded better
• Embedded reports inside the merchant-facing app

You’ll fit in if you

• Have a CS degree and enjoy SQL and ML
• Understand the math behind ML and statistics
• Write Python comfortably
• Get a little obsessed when two numbers don’t match
• Like explaining your reasoning in writing

Bonus points, none required: dbt, Dagster, Airflow, Cube, Superset or Tableau, LLM APIs, an internship or serious side project, analytical databases (Vertica, BigQuery, Snowflake).

Where you’ll be in 90 days: you’ve shipped a pipeline change, run your own evaluation of the assistant, and presented your findings to the team.