Machine Learning Data Engineer (LLM & Generative AI)

Felix Technologies, Inc.•New York, NY

1d•Remote

About The Position

Félix is seeking an innovative, highly technical Machine Learning Data Engineer to build foundational data infrastructure for its large language models (LLMs) and autonomous AI agents. The role covers the end-to-end data lifecycle, from orchestrating real-time telemetry to creating pipelines that train and refine models. The engineer will ensure models receive high-quality, diverse, well-structured data that supports continuous performance improvements. The role includes building automated pipelines for LLM fine-tuning, developing production evaluation and telemetry systems, scaling synthetic data workflows, and implementing data quality and curation practices. The engineer will also integrate data pipelines with model training infrastructure and manage data flows into vector databases for retrieval-augmented generation (RAG).

Requirements

Bachelor’s or Master’s degree in Computer Science, Data Science, Engineering, or a related technical field.
At least 5 years of professional experience in Data Engineering, Machine Learning Engineering, or MLOps.
Expert-level proficiency in Python and advanced SQL.
Hands-on experience with LLM APIs such as OpenAI or Anthropic, open-source models such as Hugging Face or Llama, and orchestration frameworks such as LangChain, LlamaIndex, or AutoGen.
Experience using prompt engineering and programmatic techniques to generate and validate high-quality synthetic data.
Strong experience with workflow orchestration tools.
Experience with scalable cloud infrastructure, specifically GCP, and distributed data processing frameworks.
Equivalent competencies in the listed areas will also be considered.

Nice To Haves

Familiarity with vector search and vector databases such as Pinecone, Milvus, Weaviate, or Qdrant.
Experience with LLM evaluation frameworks such as RAGAS, TruLens, or DeepEval, and automated LLM benchmarking suites.
Understanding of privacy-preserving techniques, including PII redaction pipelines, for sensitive data used in ML training.
Experience deploying streaming data pipelines using Kafka, Flink, or Kinesis for real-time analytics and to feed contextual data into agentic workflows and streaming RAG systems.
Bilingual English/Spanish.

Responsibilities

Design, build, and maintain scalable automated pipelines to extract, clean, and format datasets for LLM fine-tuning, including Supervised Fine-Tuning, RLHF, and DPO.
Develop telemetry and sampling systems to ingest real-time production data and evaluate and benchmark live LLM and AI-agent performance, latency, and accuracy.
Architect and scale synthetic data generation workflows using foundational models to augment training data, cover edge cases, and mitigate bias where real-world data is scarce or privacy-restricted.
Implement automated data quality checks, deduplication strategies, and toxicity filtering for model-training data.
Partner with ML Researchers and DevOps to integrate data pipelines with model-training infrastructure and support CI/CD for model updates.
Manage and optimize data flows into vector databases for RAG capabilities in agentic workflows.