Own the end-to-end solution architecture of the Daitics AI CDP: a sovereign, on-prem, telco-native Customer Data Platform built on a streaming architecture, deployed on infrastructure we run ourselves (Kubernetes, Helm), not managed cloud services. The platform processes hundreds of thousands of events per second across thirty or more source systems and delivers unified customer profiles for B2C persons and B2B organizations, accounts, sites, lines, devices and contacts. You will own the architecture across the five planes (Authoring, Control, Data, Activation, Observability) and the five layers (atomic events, tile primitives, trait values, signal filters, signal events), hold the line on architectural principles across engine boundaries, and translate them into designs engineering teams can build against.

Requirements

• 12+ years in software and data engineering, with at least 5 years owning solution or platform architecture for large-scale distributed systems

• Demonstrated ownership of end-to-end architecture for a streaming data platform on self-managed infrastructure (on-prem or private cloud) - topology, state placement, failure domains, cutover and capacity

• Deep, hands-on architectural command of Apache Flink as a stateful streaming runtime - DataStream and SQL, keyed state on RocksDB, broadcast state, Async I/O, checkpoint and savepoint discipline (core requirement)

• Proven ability to design two-tier latency architectures: a millisecond-level hot path against a minute-level durable cold path with an explicit end-to-end latency budget

• Experience designing Kafka topologies for multi-tenant platforms - topic taxonomy, partition-key selection and sizing, compacted control streams, retention classes and naming conventions

• Experience architecting schema governance across producers and consumers - Avro or equivalent through a schema registry, compatibility modes and coordinated promotion of breaking changes

• Hands-on architecture experience with lakehouse table formats (Apache Paimon, Iceberg, Delta Lake or Hudi) on S3-compatible object storage - primary-key and LSM table design, partitioning, compaction and snapshot expiry

• Experience with distributed in-memory caches (Apache Ignite, Hazelcast, Redis or similar) - partitioned versus replicated tables, affinity colocation, thin-client access and cross-consumer invalidation

• Ability to define and enforce a state-placement decision framework: broadcast state versus keyed state versus external cache versus first-seen load

• Experience architecting identity resolution at scale - deterministic and probabilistic matching, identity graph modeling, effective-dated B2B relationships, merge and split workflows and anonymous-to-known stitching

• Architecture experience with consent enforcement and data governance - configurable enforcement modes, data usage labeling, policy distribution to the runtime and designs that avoid a synchronous service call per event

• Experience architecting tokenization, hashing, masking and encryption for sensitive identifiers - key versioning and rotation (HashiCorp Vault Transit or equivalent) and audited just-in-time detokenization

• Experience designing end-to-end lineage and edit-time impact analysis, versioned artifact activation and cutover across engine boundaries with grace windows and rollback

• Experience designing multi-tenant isolation on Kubernetes, enterprise IdP federation (OIDC, Keycloak or equivalent), ABAC, NetworkPolicies, service-mesh mTLS and secrets management

• Experience defining failure domains, SLOs, RPO and RTO targets for streaming platforms, and designing replay, backfill and reconciliation paths that do not re-trigger external actions

• Practical experience with Spring Boot on Java for control-plane microservices, PostgreSQL as a definition store, and workflow orchestration (Temporal, Argo Workflows or equivalent)

• Experience with observability stacks (OpenTelemetry, Prometheus, Grafana, distributed tracing); telco experience with TM Forum SID and TMF620 is a strong advantage

• Familiarity using AI tools for design, development and debugging (Claude, Cursor, Codex)