Software Engineer (JVM)

Nexla, briefly

Nexla is the data layer for enterprise AI. We give AI apps and agents the connectivity, context, and governance to work across more than 1,000 enterprise systems in real time, and we process over a trillion records a month doing it. DoorDash, LinkedIn, Johnson & Johnson, Instacart, and LiveRamp run mission-critical data on us. Honorable Mention in the Gartner Magic Quadrant™ for Data Integration Tools, and top-rated by customers on Gartner Peer Insights four years running. Founded 2016, remote-first, headquartered in San Mateo.

Our values are short and we actually use them: Have Empathy. Be Curious. Be Intellectually Honest. Achieve Excellence. Remember to Relax.

Role

You would work on the backend systems that move data through Nexla: JVM services, the streaming runtime on Kafka, and the distributed processing underneath. Production Java every day. You would get well-scoped problems rather than a blank page, and you would own them from design through to production. Small team, so your code reaches customers quickly and you see what it does there. This role sits closest to the integration layer, where the platform meets everything outside it.

What you'll own

• Features, end to end. You shape the approach with your lead, build it, write the tests, ship it, and stay with it after it ships.
• The quality of what you write. Tests that catch real failures, and code the next person can read without a guided tour.
• Debugging your own work in production, and over time other people’s. Tracing a failure through Kafka, a distributed runtime, or a third-party API that went quiet is a skill you would build quickly here.
• Getting fluent in the domain. Schemas that change without warning, auth schemes, rate limits, retries. This is the substance of the work and it takes time to learn properly.

Roughly 80% building and shipping, 20% code review, design discussions, and learning the systems around you.

The hard problem

Our platform integrates with more than a thousand external systems, each with its own schema quirks, auth scheme, rate limits, and retry semantics. The framework problem is the N+1 one: making the thousandth integration as cheap to add as the tenth. That means absorbing the differences into the abstraction rather than special-casing them, while the data underneath keeps moving at production volume. You would not own this problem on your own, but you would be working on pieces of it from early on.

Must-Haves

• 2 to 5 years of professional software engineering, with Java as your primary language.
• Solid JVM fundamentals: collections, concurrency basics, how memory behaves, and enough to read a stack trace or a thread dump when something is wrong.
• You write tests as part of building, not after someone asks for them.
• Comfortable with SQL, and with at least one of the protocols the platform speaks: gRPC, REST, webhooks, or CDC.
• You use AI tools in your actual work and can say specifically where they help and where they mislead you.

Bonus points

• Exposure to Kafka, Spark, Flink, or a similar distributed system in production, even if you were not the one running it.
• Kubernetes, or any experience debugging something that was not running on your laptop.
• Open-source contributions, side projects, or anything you have built that you can walk us through.
• Data integration, ETL, or any work that involved cleaning up someone else’s messy data.

You'll thrive here if

• You finish things. Started and done are different, and you know which one you are at.
• You ask early when you are stuck, and you have usually tried two things first.
• You read code you did not write without complaining about it.
• You want to get a lot better quickly, and you would rather have hard feedback than polite feedback.

One thing to be clear about: our product definition is still moving as AI reshapes data integration, and the platform has to move with it. If you want a stable, well-defined surface to work on, this is not it.

How we hire

Four conversations, about two weeks end to end.

• Coding. Working through real code rather than a puzzle.
• Low-level design. How you structure a class or a small system, and why you made those calls.
• High-level design. What do you think about a system larger than anything you have built yet. We calibrate this to your experience.
• Hiring Manager. How you learn, how you own things, and how you actually use AI in your work.

On AI in the coding round. Use it the way you would on the job, ours or anyone else’s. We build AI tooling and we expect you to use AI tooling, so watching you work without it would tell us nothing useful. What we dig into is your judgment: what you delegated, what you verified, and what you threw away.

You will hear from us either way.

The practical stuff

Bengaluru, hybrid, 2 days a week in office. Compensation includes base salary and equity, set by depth and experience rather than by title. You will need to overlap with morning Pacific hours for syncs, design reviews, and collaboration with our US-based leadership and engineering teams.

Worth a look before you apply

• What we are building now.MCP Studio and Express, our two newest bets.
• Who uses us, and why.nexla.com/customers
• The product in engineer terms.docs.nexla.com
• Our CEO on why we take "Remember to Relax" seriously.Read the founder story
• Everything else.nexla.com   ·  Blog   ·  All open roles

Why Nexla? Why now?

A few large companies - DoorDash, SentinelOne, Johnson & Johnson, LinkedIn, Amex, Integrity Marketing, among them use Nexla for data integration. Connectors, runtime, transformations, scheduling, the parts of the stack where data has to move between systems reliably.

The reason this is an interesting moment to join is what the agent shift is doing to the category. Data integration used to mean "land this data in that warehouse so a human can look at it." That product is mature. What it's becoming is closer to "an agent asks a business question, and the platform figures out which data and what code and which APIs add up to a real answer." That's a much bigger problem, and most of the architecture for it hasn't been built yet by anyone.

A concrete example. A revenue team wants to ask "which enterprise customers are showing renewal risk" and get a real answer through whatever agent or app they work in. Today there is no clean way to answer that question, it requires CRM, usage, support, and billing data joined in org-specific ways, and the calling agent can't just invoke a tool that returns the right answer unless someone first establishes whether the underlying data is even capable of producing one.

That is the system we are building. A probe agent investigates the data plane whether the right fields are populated, whether freshness is adequate, whether the joins exist, whether credentials cover the required scope and returns a grounded feasibility answer. Where there are gaps, Express.dev composes the pipelines to close them. The capability is then exposed as an MCP server: "renewal risk" as a curated product, with the business logic correct and the query semantics described well enough that the calling agent uses it as intended. The MCP Gateway governs which agents can call which capabilities, with the policies, audit, and observability an enterprise control plane requires.

The bet is that ten years of connector work, an enterprise customer base, and a runtime that already moves real volume are the right foundation to define this layer from.