Junior AI Engineer

ai engineering · junior

Builds the features that call a hosted language model, grounds them in our data, and measures whether the answers hold up.

Build your own roadmap from your CV

Free. Your CV builds a skill graph, then this roadmap becomes yours.

Pythonmust have

The whole application layer, from prompt assembly to the API route, is Python.

LLM Application Developmentmust have

The product is built on a model you did not train; wiring prompts, tools and outputs into a working feature is the daily job.

OpenAI APImust have

Every feature calls a hosted model through a chat-completions style API; you must know its request shape, its limits and its bill.

Prompt Engineeringmust have

The prompt is code you cannot unit test the usual way; you write it, version it and evaluate it like code.

Embeddingsmust have

Retrieval, deduplication and search all start with turning text into vectors; you must know what they capture and what they miss.

RAGmust have

Most answers come from our documents, not the model's weights; retrieval is how an answer gets grounded.

FastAPImust have

Every model call sits behind a FastAPI route with a typed request and response; you write and test that route.

LLM Evaluationmust have

A demo that works once is not a feature; you build the eval set and the scoring that says whether a prompt change helped.

vector searchmust have

Grounding means finding the right chunk fast; you index, query and tune a vector store.

Gitmust have

Prompts, evals and routes ship through pull requests; a prompt nobody can diff is a prompt nobody can fix.

Function Callingmust have

The model acts through tools you define; you write the schema, the handler and the guard around it.

Pydanticmust have

Every model output is parsed into a typed schema before anything trusts it; Pydantic is that gate.

Dockermust have

The service ships as a container; you build it, run it locally and make it start the same way in staging.

Guardrailsmust have

User text reaches the model and the model's text reaches users; you put the checks on both sides.

asyncionice to have

SQLnice to have

LangChainnice to have

AI Agentsnice to have

LangGraphnice to have

pgvectornice to have

Hugging Facenice to have

Transformersnice to have

Fine-tuningnice to have

LLM Orchestrationnice to have

pytestnice to have

LLM Cost Optimizationnice to have

LLMOpsnice to have

Observabilitynice to have

API Securitynice to have

Human-in-the-Loopnice to have

Responsible AInice to have