Own your future:
Our culture isn't something people join, it's something they build and shape. We believe that every person deserves to be heard and empowered. If you're on the fence about whether you're a fit, we say go for it. Let's build something great together.
Must Haves:
* 5+ years of professional Python development using modern practices - type hints, async/await, Pydantic, and a focus on clean, testable code.
* LLM application engineering: building and shipping LLM-based features (OpenAI, Anthropic, Bedrock, Vertex, etc.).
* Prompt engineering: structured outputs, few-shot prompting, system prompts, tool/function calling.
* Document & unstructured data processing: PDFs, DOCX, HTML, images (parsing, chunking, extraction).
* NLP fundamentals: embeddings, tokenization, similarity, classification, entity extraction.
* Evaluation & testing of LLM systems: golden sets, regression tests, LLM-as-judge, A/B testing, drift detection.
* PostgreSQL + ORM (preferably SQLAlchemy).
* Configuration-driven design: YAML/JSON-schema-based systems instead of hardcoding logic
* Testing discipline: pytest, test-driven mindset, writing tests alongside code.
* Production engineering basics: logging, tracing, prompt/versioning awareness, reproducibility
* Real production LLM experience: at least one shipped system using LLM APIs.
Nice to Have:
* NumPy / Pandas / scikit-learn (for metrics, validation, analysis).
* LangChain / LangGraph / LangSmith or similar orchestration tools.
* Vector databases / RAG systems.
* Experience in regulated industries (healthcare, pharma, finance).
* Information extraction/knowledge graphs.
Key Responsibilities:
* Design, develop, and optimize LLM-powered applications and AI features across providers such as OpenAI, Anthropic, Bedrock, Vertex AI, and open-source models.
* Build and improve prompt engineering solutions, including system prompts, structured outputs, tool/function calling, and retrieval-augmented generation (RAG).
* Develop solutions for processing and understanding unstructured data, including PDFs, DOCX, HTML, images, tables, and forms.
* Apply NLP techniques such as embeddings, semantic search, classification, and entity extraction to solve business challenges.
* Establish AI evaluation frameworks using regression testing, golden datasets, A/B testing, and drift detection to ensure measurable quality.
* Optimize AI systems for cost, latency, and performance through model routing, caching, batching, and fallback strategies.
* Design configuration-driven AI pipelines and structured output validation using YAML/JSON schemas and Pydantic models.
* Implement data quality checks, hallucination prevention, PII filtering, and AI safety guardrails.
* Maintain production-grade engineering practices, including logging, tracing, prompt versioning, observability, and reproducible workflows.
* Collaborate with cross-functional teams to deliver scalable, reliable, and high-quality AI solutions.