About the role
- Build and productionize AI/ML systems, including large language model (LLM) applications, retrieval-augmented generation (RAG) pipelines, and model-serving infrastructure.
- Design and implement RAG architectures: chunking and embedding strategies, vector store selection and tuning, retrieval and re-ranking, and grounding responses to reduce hallucination.
- Fine-tune, adapt, and evaluate models using PyTorch, TensorFlow, and the Hugging Face ecosystem (Transformers, Datasets, Tokenizers, PEFT/LoRA).
- Develop robust evaluation frameworks — offline benchmarks, human-in-the-loop review, and online metrics — to measure quality, latency, and cost.
- Optimize inference for performance and cost (quantization, batching, caching, GPU utilization).
- Write clean, well-tested, maintainable code and contribute to shared libraries and services.
- Collaborate with product and design to scope AI features, and communicate trade-offs clearly to technical and non-technical stakeholders.
- Monitor deployed systems for drift, regressions, and reliability, and iterate based on real-world feedback.
- Bachelor’s or Master’s degree in Computer Science, Machine Learning, a related field, or equivalent practical experience.
- AI Frameworks: Working knowledge of PyTorch, TensorFlow, and the Hugging Face ecosystem.
- RAG: Hands-on experience building retrieval-augmented generation systems — embeddings, vector databases (e.g., FAISS, Pinecone, Weaviate, pgvector), and retrieval pipeline design.
- LLMs: Practical experience working with large language models, including prompting, fine-tuning, and evaluation.
- Strong programming skills in Python, with solid software engineering fundamentals (version control, testing, code review).
- Experience deploying models or ML services to production (APIs, containers, cloud environments).
- Familiarity with the ML lifecycle: data preparation, experimentation, evaluation, deployment, and monitoring.
- Ability to read research papers and translate ideas into working implementations.
- Experience with orchestration frameworks such as LangChain, LlamaIndex, or equivalent.
- Familiarity with agentic systems, tool use, and function calling.
- Experience with inference optimization (vLLM, TensorRT, ONNX, quantization) and GPU workloads.
- Exposure to MLOps tooling (MLflow, Weights & Biases, Kubeflow) and CI/CD for ML.
- Cloud experience (AWS, GCP, or Azure) and infrastructure-as-code.
- Knowledge of responsible AI practices: evaluation for safety, bias, and reliability.
- Open-source contributions or a portfolio of applied AI projects.
- Competitive salary and equity.
- [Health, retirement, and other benefits].
- Budget for compute, learning, and conferences.
- A collaborative team building AI products that reach real users.
- [Flexible work arrangements / other perks].