Distributed data and infrastructure

Replicated partitioned commit logs, zero-copy socket transfers, and end-to-end consumer offsets.

Continuous stream analytics, out-of-order event time watermarks, and Chandy-Lamport state checkpoints.

Formal consensus via Raft state machines, leadership heartbeats, etcd leases, and monotonic fencing tokens.

Decentralized Dynamo rings, sloppy quorums, Merkle anti-entropy, and Multi-Raft transactional SQL engines.

Reed-Solomon erasure coding over Galois fields, chunk placement algorithms, and petabyte-scale durability.

Distributed inverted indexes, segment merge policies, and SIMD columnar vector execution.

Automated node remediation, cluster drain runbooks, live partition rebalancing, and data repair.

Architect and implement a production-grade distributed key-value store from scratch: LSM storage, Raft, and consistent hash routing.