Principal Software Engineer, AI Platform Engineering

Saviynt

Internet Software

Milpitas, CA - USA

Entry Level

Experteer Overview

In this role you shape the AI platform's data architecture, governing training data flow and ensuring tenant isolation and privacy. You will set standards used by ML engineers and scientists, influencing end-to-end data lifecycle from source to model. You’ll own infrastructure for data lakes, pipelines, schema management, and multi-tenant safety, enabling scalable, compliant AI workloads. This position offers impact across a large-scale SaaS platform and collaboration with cross-functional teams to advance secure, reliable AI capabilities.

Compensation / Benefits

competitive compensation
growth opportunities
reliable, enterprise-grade platform
collaborative, reliability-focused culture

Responsibilities

Define and govern data lake architecture on GCS (raw, silver, gold; CMEK; lifecycle)
Build and maintain batch pipelines (Spark on Dataproc; Iceberg maintenance; daily S3GCS sync)
Develop streaming pipelines (Apache Beam on Dataflow; exactly-once semantics; PII gating)
Manage schema registry (Avro/Protobuf versioning; migration playbooks)
Oversee orchestration (Flyte; domain isolation; retry policies; DataCatalog memoization)
Enforce multi-tenant data architecture (per-tenant isolation; quotas; contamination checks)
Develop Data Anonymizer and Data Labeler microservices for PII stripping and labeling
Operate feature store (offline/online; consistency; PIT joins) and vector store (Pgvector; Qdrant)
Design RAG data pipelines (embedding generation; upsert; freshness SLAs)
Expose data platform services via API (gRPC/HTTPS; mTLS)
Create synthetic data pipelines for dev/staging
Implement data quality gates (Great Expectations/dbt) as deployment-time checks

Key requirements

8+ years of data engineering at production scale
Proven impact leading platform-wide standards or major migrations
Data lake ownership and end-to-end operation experience
Deep Spark (PySpark/Scala) expertise; Iceberg/Delta Lake maintenance
Hands-on Beam/Dataflow experience (windowing; exactly-once)
Schema registry experience (Protobuf/Avro compatibility; migrations)
Production orchestration at scale (Flyte; Kubeflow Pipelines; Airflow; Prefect)
Multi-tenant architecture with strict isolation
Feature store operations (Feast or Tecton); point-in-time joins; online/offline consistency
Vector databases in production (Pgvector, Qdrant); embedding upserts; index strategies
RAG fundamentals: chunking, embeddings, retrieval quality; context freshness
API transport (gRPC, HTTPS/mTLS); proto contracts
Bachelor in Computer Science
equivalent practical experience

Strong cross-team collaboration Strategic thinking with governance in mind Proactive problem solving and ownership PySpark/Scala Spark on Dataproc; Iceberg/Delta Lake Beam/Dataflow; exactly-once; windowing

Description

In this role you shape the AI platform's data architecture, governing training data flow and ensuring tenant isolation and privacy. You will…

1M+ top positions worldwide with salary benchmarks
Be discreetly found and contacted by headhunters
Exclusively for senior-level professionals and executives

Already a member?