Overview

We’re hiring a Software Engineer – AI & Data Science to build and ship production AI systems across real client and internal use cases.

Responsibilities

• Build AI-powered applications using LLM APIs (OpenAI, Anthropic, Google, and open-source models)
• Build RAG pipelines, and tune retrieval quality using embeddings, chunking strategies, and reranking
• Design and run evaluations for LLM and ML systems: define metrics, build test sets, and measure quality
• Develop AI agents, tool-calling workflows, and multi-step LLM systems
• Build and maintain data pipelines that prepare, clean, and transform data for AI and analytics use cases
• Train, fine-tune, or adapt models where off-the-shelf approaches fall short
• Analyze data and model outputs to find failure modes and drive improvements
• Ship your work as production code, including Python services and APIs, working with the engineering team
• Work closely with product, founders, and client teams to turn business problems into data and AI solutions

Qualifications

• 2 to 5 years into your professional career in data science, ML, or AI engineering
• A data scientist who writes production-quality code, not just notebooks
• Someone who measures before claiming something works
• Comfortable picking up new models, frameworks, and tools quickly
• Someone who enjoys owning a problem end to end, from data to deployed system
• Strong Python, including pandas, NumPy, and writing clean, tested, production-ready code
• Solid SQL and experience working with real-world, messy data
• Strong grasp of ML fundamentals: statistics, model evaluation, and experimentation
• Hands-on experience with LLM APIs, including prompt engineering, structured outputs, and tool calling
• Experience building RAG pipelines or AI agents (LangChain, LlamaIndex, LangGraph, or similar)
• Experience designing evaluations for ML or LLM systems
• Strong problem-solving and communication skills

Nice to have

• Fine-tuning LLMs or training ML models (PyTorch, Hugging Face, scikit-learn)
• Vector databases: Pinecone, pgvector, Weaviate, Qdrant, Chroma, or similar
• LLM observability and evaluation tools (Langfuse, LangSmith, Braintrust, or similar)
• Building APIs with FastAPI or Flask
• Data tooling: dbt, Airflow, Spark, or similar
• Docker, and cloud (AWS, GCP, or Azure)