We are looking for a hands-on Senior Data Engineer to join the Data Platform team and own the Data Movement platform: the systems that bring data into MoMo's data platform and keep it in sync across our storage and compute environments. This role covers two responsibilities. The first is data ingestion: batch and real-time pipelines that land data from a wide range of internal and external sources into the platform, reliably and at scale. The second is data replication: moving and synchronizing data between BigQuery and MoMo's Lakehouse infrastructure, across GCP and on-premise environments. Data movement is the layer everything else depends on. Every warehouse model, dashboard, report, and AI application at MoMo is only as trustworthy as the pipelines underneath it. You will be accountable for making that layer boring in the best sense: consistent, available, secure, observable, and predictable, with clear guarantees that downstream teams can build on rather than work around. Mô tả công việcData ingestion (batch and real-time)Design, build, and operate ingestion pipelines that bring data into the platform from diverse sources, including operational databases, application events and message streams, internal services and APIs, files, and third-party partner feeds. Build the real-time ingestion path on a streaming stack (Kafka, Flink or Spark Structured Streaming, or equivalent), covering schema handling, exactly-once or effectively-once delivery, late and out-of-order events, and backpressure. Build the batch ingestion path with reliable scheduling, incremental extraction, watermarking, idempotent loads, and safe re-runs and backfills. Standardize ingestion as a repeatable, configuration-driven capability rather than one bespoke pipeline per source, so onboarding a new source is a routine operation with a known cost and timeline. Define and enforce the contract at the ingestion boundary: schema registry and schema evolution rules, data typing, deduplication, ordering guarantees, and how upstream breaking changes are detected before they reach consumers. Data replication between BigQuery and the LakehouseDesign and operate replication between BigQuery and MoMo's Lakehouse infrastructure (Iceberg on GCP and on-premise), in both directions and across environments. Handle the hard parts of cross-platform replication: type and schema mapping, partitioning strategy, incremental and CDC-based sync, snapshot and full-refresh strategy, large historical backfills, and cost and throughput optimization. Build reconciliation and validation as a first-class part of the platform: row and aggregate level checks, drift detection between source and target, automated alerting, and a documented remediation path when a mismatch is found. Support platform migration work by making replication a safe bridge during transition, so workloads can run in parallel across BigQuery and the Lakehouse and cut over without data loss or downtime. Consistency, availability, and securityDefine the guarantees the platform provides per pipeline (delivery semantics, freshness, completeness, ordering) and make them explicit, measurable, and monitored rather than assumed. Own availability and reliability: SLA and SLO definition, freshness and lag monitoring, failure detection, retry and dead-letter handling, replay procedures, capacity planning, on-call runbooks, and post-incident follow-through. Build the observability layer for data movement: pipeline health, throughput, lag, error rates, cost per pipeline, and lineage from source to target, so problems are found by monitoring rather than by a downstream user. Implement security and privacy controls in transit and at rest: authentication and authorization, secrets management, encryption, network boundaries, access control on replicated datasets, audit logging, and handling of PII in line with data residency and privacy requirements. Automate the platform through infrastructure-as-code, CI/CD, and self-service tooling, and write the documentation and runbooks that let other engineers operate and extend it. Yêu cầu công việc4+ years of experience in Data Engineering or Data Platform Engineering, with proven hands-on delivery and operation of production data pipelines at scale. Strong programming skills in Python, Java, or Scala, with solid software engineering practices: Git workflow, code review, testing, and CI/CD. Hands-on experience building real-time / streaming pipelines with Kafka and a stream processing framework such as Flink, Spark Structured Streaming, or equivalent, including delivery semantics and state management. Hands-on experience with batch ingestion and CDC tooling and patterns (Debezium, Airbyte, Sqoop-class tools, DataStream, or in-house equivalents) against relational and NoSQL sources. Hands-on experience with Apache Spark and distributed data processing, including performance tuning and cost optimization. Working knowledge of BigQuery and an open Lakehouse stack: Apache Iceberg (or Delta/Hudi), object storage, Hive Metastore or an equivalent catalog. Experience with workflow orchestration in production (Airflow or equivalent), including SLA ownership, backfills, and incident handling. Strong grasp of distributed systems fundamentals: consistency models, delivery guarantees, idempotency, partitioning, ordering, and failure modes, with the ability to reason about correctness under partial failure. Experience operating platforms in production with an SRE mindset: monitoring and alerting, on-call, incident response, root cause analysis, and capacity planning. Familiarity with containers and infrastructure-as-code (Kubernetes, Docker, Terraform) and with cloud and on-premise networking basics. Experience with data security and privacy controls (encryption, access control, secrets management, PII handling, audit) is a strong plus. Experience with hybrid cloud and on-premise data movement, or with cross-cloud replication, is a strong plus. Ownership mindset, comfort with production accountability, and the ability to work with both platform engineers and downstream data consumers.