Requirements

Expected technologies

Kubernetes

Helm

Terraform

Redis

Prometheus

Grafana

Aerospike

GitHub

GitLab

Docker

Python

SQL

Spark

Flink

Kafka

TensorFlow

PyTorch

Optional technologies

Seldon

Triton

ONNX

TensorRT

AWS

IAM

S3

Lambda

Step Functions

RDS

DynamoDB

EMR

Snowflake

Snowpark

Rails

React

Our requirements

• Degree in Computer Science or related field (or equivalent experience).
• At least 3 years of previous leadership experience
• Kubernetes in production (EKS/ECS): deployments, secrets/config, Helm/Terraform.
• Strong experience building low-latency, high-throughput microservices.
• ML serving systems: architecture, real-time monitoring/alerting (Prometheus/Grafana), distributed caching (Redis/Aerospike).
• CI/CD & packaging: GitHub/GitLab, Docker, Python packaging/PyPI; REST APIs and SDK/library design.
• Familiarity with ETL/big-data and mainstream ML libraries (Spark/Flink, Kafka, TensorFlow/PyTorch).
• Proficient in Python and SQL, solid algorithms/data structures, and strong debugging in production.
• Excellent communication and cross-functional collaboration.

Optional

• 3+ years building low-latency microservices at scale.
• Experience with ML serving frameworks (Seldon, Triton, ONNX/TensorRT).
• AdTech/RTB experience.
• Deep AWS knowledge (IAM, S3, Lambda/Step Functions, RDS, DynamoDB, EMR).
• Snowflake/Snowpark; Redis/Aerospike at scale.
• Full-stack exposure (Rails/React) for internal tooling.

Your responsibilities

• Lead & deliver: Own roadmap and execution for ML Platform and ML Serving; plan quarterly, manage risks, and ship reliably.
• Platform at scale: Design the next-gen ML platform to support thousands of concurrent training pipelines and high-volume batch predictions.
• Ultra-low-latency serving: Build and evolve real-time model serving for high throughput, low latency, and five-nines availability.
• Production excellence: Establish CI/CD for models (packaging, versioning, canary/blue-green, rollback), real-time monitoring/alerting, and SLOs.
• Experiment & accelerate: Prototype new platform/serving ideas; reduce model lead time from research to production.
• Collaborate globally: Partner with model engineers, data, product, and infra to deliver ML features that move the business.
• Mentor & grow: Provide technical guidance and raise the bar on engineering quality.