Data / ML Platform Engineer

Kela is a defense technology startup pioneering next-generation command & control and data infrastructure systems. We empower defense organizations with robust, scalable, and secure platforms that integrate diverse sensors and AI to deliver real-time battlefield intelligence. Backed by leading investors, Kela combines Israeli innovation with global mission impact.
The Platform as a Service team is Kela's customer-facing engineering function: it delivers customer projects on the Kela platform and turns what it learns into reusable products.
We are looking for a Data / ML Platform Engineer to sit between our data platform and our customer-facing work. You will build the data and ML infrastructure and work directly with customer algorithm teams to drive their projects, including on-site and forward-deployed when the work needs it. You are not only a builder of internal tooling, and you are not only a customer-facing integrator. You do both, with a strong data foundation.
The role is hybrid by design. It is expected to grow toward either the data platform side or the customer-facing side, based on your preference and what the team needs. You will also seed new data products from an early idea, then carry them into real customer projects.
You will be a core member Kela's PaaS team, while working closely with our core software, algorithms, and DevOps teams.
Responsibilities:

Support customer algorithm teams across the full ML lifecycle: data annotation, training, evaluation, and observability, covering both the platform infrastructure and the hands-on project work
Work directly with customer teams to understand their data and ML needs, unblock them, and turn what you learn into platform improvements
Seed new data products from scratch, and take them forward by deploying and adapting them in customer projects
Build and operate data and ML pipelines for multimodal sensor data (video, radar, RF, telemetry)
Set up and maintain the data environments, model onboarding flows, and evaluation workflows that customer-side research teams rely on
Partner with algorithms, product, and DevOps teams to establish data and MLOps standards across the company

Requirements:

At least 3 years of experience in data engineering, MLOps, or backend engineering, building production systems
Strong proficiency in Python, with experience writing production-grade code and pipelines
Strong SQL skills and solid data modeling fundamentals
Hands-on experience with Kubernetes and containerized deployments
Experience with ML workflows in practice: training, evaluation, experiment tracking, and model deployment
Comfortable working directly with customers or technical stakeholders, and able to turn an unclear request into a working solution
Ability to work independently, and to move between building infrastructure and delivering a project
Demonstrated full-stack “builder” mentality through professional or personal projects

Preferred Qualifications:

An algorithms background: able to build and run evaluations that compare models (for example, re-identification models)
Experience with a semantic layer (dbt Semantic Layer, Snowflake Semantic Views) or ontology technologies (for example, Palantir)
Experience with MLOps frameworks (for example, MLflow, Kubeflow, Airflow orchestration)
Experience with data lake architectures (Iceberg, Delta) and query engines (Snowflake, Athena, BigQuery)
Familiarity with ELT/ETL pipeline design (Airflow, dbt)
Exposure to multimodal sensor data (EO/IR, radar, RF, geospatial, IMU, LiDAR)
Background in defense, aerospace, or other mission-critical systems