PhonePe
About this role
PhonePe runs payments infrastructure at a scale where standard tools stop behaving normally, and this role lives at exactly that layer. It is one of the most specialised listings in today's edition: you would own PhonePe's internal Apache Kafka distribution, covering branch management, patch curation, release cutting and upgrade planning, while contributing back upstream through bug fixes, KIPs, code reviews and design discussions. The day to day work is diagnosing deep broker and controller problems, meaning replication, partition reassignment, KRaft consensus, log storage, in sync replica management and protocol handling, then building custom patches for production needs that are not suitable to upstream, such as performance tuning for PhonePe's own workload profile, observability hooks and operational safety valves. You would set how the team tests, benchmarks and validates patches before rollout, mentor a supporting engineer, act as the internal authority other platform teams consult on Kafka behaviour and capacity planning, and serve as the interface to audit and compliance teams.
Who this is for
What the posting requires
- 8+ years in systems or infrastructure software engineering, with a meaningful portion on Kafka or comparable distributed systems such as Pulsar, BookKeeper, ZooKeeper, etcd or CockroachDB
- Deep familiarity with the Kafka codebase. The posting draws a sharp line here: it wants people who have read, modified and debugged the broker, controller or client internals, not people who have operated Kafka
- A track record of upstream contribution: committed patches, participation in KIP discussions, reviewed pull requests or equivalent engagement with the Apache Kafka community. Committer or PMC status is strongly preferred but not strictly required
- Strong distributed systems fundamentals understood as engineering tradeoffs you have navigated: consensus protocols, replication, exactly once semantics, log structured storage and distributed state machines
- Systems level programming comfort: JVM internals including garbage collection tuning, memory layout and JIT behaviour; operating system I/O including page cache, fsync semantics and zero copy; and networking including TCP tuning and socket buffers
- Experience maintaining a fork or distribution, and managing the tension between carrying local patches and staying mergeable with upstream
Counts as a bonus, not a requirement
- Apache Kafka committer or PMC membership
- Experience with KRaft, the ZooKeeper free controller
- Having authored or co-authored a KIP
- Kafka Streams or Kafka Connect internals, rather than just usage
- Familiarity with Kafka's own test infrastructure including system tests, Trogdor and ducktape
What you would actually be doing
- Owning the internal Kafka distribution: branches, patch curation, releases and upgrade planning
- Contributing upstream to Apache Kafka
- Diagnosing and fixing deep broker and controller issues across replication, partition reassignment, KRaft consensus, log storage and protocol handling
- Building custom patches for production needs unsuited to upstream
- Establishing how the team tests, benchmarks and validates patches before rollout
- Mentoring a supporting engineer and growing the team's capability
- Advising platform teams on configuration, capacity planning and architecture, and acting as the primary interface for audit and compliance
Location and working pattern
Bangalore. Staff at PhonePe is a senior individual contributor track.
Also worth knowing
PhonePe's listed benefits include comprehensive insurance, an onsite medical centre, parental and adoption support, relocation help, provident fund and gratuity, higher education assistance and a car lease policy.
A good fit if
You have gone deep on infrastructure internals rather than broad across product features, and you want one of the few Indian roles that will pay you to work on an open source distributed system at source level.
Think twice if
You have operated Kafka but never modified it. The posting makes that distinction explicitly, and the interview will go straight there.
- 8+ years in systems or infrastructure software engineering, with a meaningful portion on Kafka or comparable distributed systems such as Pulsar, BookKeeper, ZooKeeper, etcd or CockroachDB
- Deep familiarity with the Kafka codebase. The posting draws a sharp line here: it wants people who have read, modified and debugged the broker, controller or client internals, not people who have operated Kafka
- A track record of upstream contribution: committed patches, participation in KIP discussions, reviewed pull requests or equivalent engagement with the Apache Kafka community. Committer or PMC status is strongly preferred but not strictly required
- Strong distributed systems fundamentals understood as engineering tradeoffs you have navigated: consensus protocols, replication, exactly once semantics, log structured storage and distributed state machines
- Systems level programming comfort: JVM internals including garbage collection tuning, memory layout and JIT behaviour; operating system I/O including page cache, fsync semantics and zero copy; and networking including TCP tuning and socket buffers
- Experience maintaining a fork or distribution, and managing the tension between carrying local patches and staying mergeable with upstream
Counts as a bonus, not a requirement
- Apache Kafka committer or PMC membership
- Experience with KRaft, the ZooKeeper free controller
- Having authored or co-authored a KIP
- Kafka Streams or Kafka Connect internals, rather than just usage
- Familiarity with Kafka's own test infrastructure including system tests, Trogdor and ducktape
What you would actually be doing
- Owning the internal Kafka distribution: branches, patch curation, releases and upgrade planning
- Contributing upstream to Apache Kafka
- Diagnosing and fixing deep broker and controller issues across replication, partition reassignment, KRaft consensus, log storage and protocol handling
- Building custom patches for production needs unsuited to upstream
- Establishing how the team tests, benchmarks and validates patches before rollout
- Mentoring a supporting engineer and growing the team's capability
- Advising platform teams on configuration, capacity planning and architecture, and acting as the primary interface for audit and compliance
Location and working pattern
Bangalore. Staff at PhonePe is a senior individual contributor track.
Also worth knowing
PhonePe's listed benefits include comprehensive insurance, an onsite medical centre, parental and adoption support, relocation help, provident fund and gratuity, higher education assistance and a car lease policy.
A good fit if
You have gone deep on infrastructure internals rather than broad across product features, and you want one of the few Indian roles that will pay you to work on an open source distributed system at source level.
Think twice if
You have operated Kafka but never modified it. The posting makes that distinction explicitly, and the interview will go straight there.
Apply on company site
Opens job-boards.greenhouse.io, the employer's own application page. Applying is always free.