Senior ML Encoder Lead | Customer Embeddings Architect

Planet Pharma; South San Francisco (CA); Hybrid; USD 110,000 - 146,000; Full time

Job summary

Planet Pharma is seeking an ML Encoder Lead (Senior Contractor) focused on Customer Representation Learning and Encoder Development. Build a shared customer representation, train and evaluate the encoder, and provide evidence for continuation.

You will define modeling objectives, evaluation design, and write production code to deploy scalable solutions. The role demands hands-on work, ownership of modeling objectives, and credible presentation to stakeholders; Bay Area based with 3 days in

Qualifications

• Personally trained an encoder or embedding model with a designed pretraining objective.
• Deep expertise in self-supervised/contrastive learning and sequence modeling.
• Experience with long, sparse, longitudinal event data (transactions, journeys, etc.).
• Ability to build inductive representations for entities with limited history.
• Rigorous evaluation: time-based splits, leakage checks, cold-start tests, baselines.
• Assess embedding usefulness for downstream tasks and report uncertainty.
• Strong Python engineering with PyTorch or JAX and cloud-scale training.

Responsibilities

• Design the pretraining objective and evaluation strategy for the encoder.
• Train and evaluate the encoder on longitudinal customer data.
• Produce evidence to decide whether to continue the approach.
• Write production code and manage data contracts, training pipelines and monitoring.

Skills

Encoder training

Representation learning

Python engineering

PyTorch

JAX

Distributed data processing

Tools

PyTorch

JAX

SQL

Cloud training

Job description

Planet Pharma is seeking an ML Encoder Lead (Senior Contractor) focused on Customer Representation Learning and Encoder Development. Build a shared customer representation, train and evaluate the encoder, and provide evidence for continuation.

You will define modeling objectives, evaluation design, and write production code to deploy scalable solutions. The role demands hands-on work, ownership of modeling objectives, and credible presentation to stakeholders; Bay Area based with 3 days in