Prophecy

Machine Learning Research Engineer

Prophecy

full-time

Posted on:

Location: California • 🇺🇸 United States

Visit company website
✨ AI Apply
Apply

Salary

💰 $250,000 - $350,000 per year

Job Level

Junior

Tech Stack

AWSCloudJavaKubernetesMicroservicesNeo4jPythonPyTorchScalaSparkSQL

About the role

  • Work across the full ML lifecycle from data acquisition to training to productization using PyTorch and open-source transformers
  • Implement advanced ML/LLM techniques including fine-tuning (SFT, DPO, RLHF), distillation, quantization/pruning, and model optimization
  • Build and integrate backend pipelines and microservices in public cloud environments (AWS, Kubernetes), collaborating with engineering teams
  • Take ideas to production with end-to-end ownership of projects and productization of ML models
  • Optimize models and systems for low-latency, high-throughput production use
  • Collaborate with cross-functional teams to deliver scalable ML-driven features for Prophecy's data integration platform

Requirements

  • Advanced ML/LLM knowledge: hands-on fine-tuning (SFT, DPO, RLHF), distillation, quantization/pruning for model optimization
  • Full ML lifecycle: from data acquisition to training to productization with PyTorch and open-source transformers (e.g., HuggingFace)
  • Fluency in Python for dataset/model processing
  • Experience with backend pipelines/microservices software development in public cloud: AWS, k8s, java, scala
  • Builder mentality: Experience taking ideas to production
  • ML/LLM work in code generation (e.g., Codex, text-to-SQL), semantic extraction, or knowledge graphs (e.g., Neo4j, Neptune) (nice-to-have)
  • Experience with big-data engines like Spark (nice-to-have)
  • Compiler development for languages like SQL, Python, or Scala (nice-to-have)
  • Optimization of ML models for low-latency, high-throughput production use (nice-to-have)
  • Contributions to open-source AI/ML projects (e.g., Hugging Face, PyTorch) (nice-to-have)
  • Expertise in retrieval systems or vector databases (e.g., Pinecone, Weaviate) (nice-to-have)
  • Skill in evaluating tech and driving build/buy decisions (nice-to-have)
  • Seniority: Jr to Staff Level
  • Experience: 1+ years in industry