SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Nexla is the data layer for enterprise AI, enabling AI apps and agents to connect, contextualize, and govern data across 1,000+ enterprise systems in real time. The company processes over a trillion records monthly and serves customers including DoorDash, LinkedIn, Johnson & Johnson, Instacart, and LiveRamp. Founded in 2016, Nexla is remote-first with headquarters in San Mateo.
In this role, you will work on backend systems that move data through Nexla's platform: JVM services, the streaming runtime on Kafka, and distributed processing infrastructure. You will own production Java systems end-to-end—from problem definition through design, implementation, deployment, and on-call support. You will work in a small, focused team where ownership is real rather than nominal.
Key responsibilities include:
• Core framework features: schema evolution, pagination, rate limiting, circuit breaking, automated retries, and self-recovery mechanisms that keep pipelines running without manual intervention. You will design and ship SDKs and interfaces that other teams build on.
• Streaming and processing in production: Kafka, Spark, and Flink workloads, including checkpointing, backpressure, and failure recovery. You operate what you build.
• Design decisions within your scope and a real voice in broader architectural choices. Multi-tenancy, concurrency, and latency considerations appear throughout; you are expected to have informed views on tradeoffs rather than only implementing others' decisions.
• Hard debugging: tracing failures across Kafka clusters, distributed runtimes, and flaky third-party APIs, then building guardrails to prevent recurrence.
The core technical challenge: Nexla integrates with 1,000+ external systems, each with unique schema quirks, auth schemes, rate limits, and retry semantics. The framework problem is making the thousandth integration as cheap to add as the tenth—absorbing differences into abstraction rather than special-casing them, while data moves at production volume.
Work allocation is roughly 70% building and shipping, 15% design and technical discussion, 15% code review, RFCs, runbooks, and documentation.
The role is hybrid in Bengaluru (2 days per week in office) with required overlap on morning Pacific hours for syncs, design reviews, and collaboration with US-based leadership and engineering teams.
REQUIREMENTS:
Must-haves:
• 5–8 years of software engineering, with meaningful experience on backend platforms, services, or SDKs (not only end-user applications)
• Strong Java: JVM memory management, concurrency models, non-blocking I/O; ability to debug thread starvation from heap dumps
• Hands-on production experience with distributed systems: Kafka, Spark, Flink, or similar (you have run them, not only built against them)
• Comfort with modern data stack and protocols: Snowflake, Databricks, gRPC, GraphQL, webhooks, CDC
• Ownership of production systems and understanding of good operational discipline
• Active use of AI tools in your work with specific knowledge of where they help and where they mislead
Bonus points:
• Open-source contributions to data projects (Airbyte, Flink, Debezium, Arrow)
• Experience at iPaaS or ELT companies (MuleSoft, Fivetran, Informatica, Airbyte)
• Deep Kubernetes and cloud-native experience, including Strimzi for Kafka
• Security architecture expertise: OAuth 2.0, mTLS, encrypted credential management at scale
