Chassis PT GCP Cloud Platform Engineer MBSE (Tamil Nadu)

Role & responsibilities : We are looking for a GCP Cloud Platform Engineer with solid hands-on experience in Google Cloud Platform, Pub/Sub, Vertex AI, Cloud Run, Cloud Build, Python and SQL. Were Hiring | GCP Cloud Platform Engineer MBSE Job ID: MBSE Global Cloud Platform Requirement 1 Experience: 46 Years Location : Chennai CTC: As Per Right Candidature Domain: Automotive / Engineering / MBSE Role: GCP Cloud Platform Engineer The candidate will work on cloud migration, distributed job orchestration, AI/ML training pipelines and scalable cloud infrastructure supporting engineering simulation workloads. 1. Distributed Systems & Messaging Concepts This is the most critical conceptual domain knowledge, separate from just "knowing the Pub/Sub API": At-least-once delivery semantics understanding that messages can be delivered more than once, and why your worker code needs to be idempotent (i.e., safe to run twice without corrupting data) Ack deadlines and lease extension since FEA jobs can run long, the candidate needs to understand how to extend acknowledgment deadlines for long-running tasks (otherwise Pub/Sub will think the worker died and redeliver the job to someone else) Dead-letter queues knowing how to design a "graveyard" queue for jobs that fail repeatedly, so bad jobs don't loop forever Ordering guarantees (or lack thereof) Pub/Sub doesn't guarantee message order by default, so if job sequencing matters at all, they need to know how ordering keys work 2. GCP-Specific Infrastructure Knowledge IAM & service accounts Pub/Sub and Vertex AI both rely heavily on fine-grained permissions; candidate should know how to scope service accounts securely (principle of least privilege) rather than granting broad access Terraform provider quirks for GCP the google Terraform provider has specific resource types for Pub/Sub topics/subscriptions and Vertex AI that behave differently than generic cloud resources Networking basics (VPC, firewall rules) especially if worker workstations are on-prem/local and need secure connectivity to GCP resources 3. Container & CI/CD Domain Knowledge Docker fundamentals building lean, reproducible training containers Cloud Build pipeline syntax and triggers knowing how to wire up automated builds tied to Git pushes or version tags Vertex AI-specific packaging conventions Vertex AI Custom Jobs expect containers to follow certain conventions (entrypoints, environment variables like AIP_MODEL_DIR, etc.) 4. Cross-Language Integration Knowledge (the unusual one) MATLAB-to-GCP interoperability this is a genuinely rare skill. Most cloud engineers have never touched MATLAB. Candidate should know either: MATLAB's limited native GCP support, or How to call GCP REST APIs / gRPC directly from MATLAB when no SDK exists This is worth flagging as a "nice to have but rare" you may need to accept a candidate who is strong on the Python/Pub/Sub side and pair them with someone who knows MATLAB, rather than expecting one person to have both. 5. Light Domain Knowledge of FEA / Engineering Simulation Workflows (Optional but valuable) They don't need to be an FEA engineer, but understanding basics like: Why FEA jobs can run for a long time and vary a lot in duration (which affects ack-deadline and autoscaling design) The nature of "design info" as structured/large input files (which affects how you'd design message payloads likely passing references/pointers to storage rather than raw data through Pub/Sub, since Pub/Sub has message size limits) This helps them make smarter architecture decisions (e.g., knowing Pub/Sub messages are capped at 10MB, so large FEA design files should live in Cloud Storage with just a reference/URL passed in the message) 6. ML/AI Training Pipeline Knowledge Basic ML training workflow concepts checkpointing, distributed training, GPU utilization to make good decisions about Vertex AI machine types and scaling Vertex AI's job specification format different from Cloud Run Jobs' YAML/config structure, so familiarity with translating one into the other is valuable Candidate has to undergo scripting test in SQL during interview. Preferred candidate profile : Pub/Sub Migration (Job Orchestration) Redesign our current BigQuery-based job orchestration system (jobs table with lease/complete state transitions, driven by frequent small DML updates) into a Pub/Sub-based work-queue architecture Rework existing Terraform-managed infrastructure to provision Pub/Sub topics, subscriptions, IAM, and dead-letter handling Update local worker client code (Python and MATLAB) to consume from Pub/Sub instead of polling BigQuery Design for idempotency and safe handling of redelivery/ack-deadline edge cases given long-running FEA jobs Address MATLAB's lack of native Pub/Sub SDK support (e.g., via REST/gRPC integration) Vertex AI Migration (AI Training Pipeline) Migrate an .