Senior Analytics Engineer

About the company

Nala is a NYC-based developer of global stablecoin payments infrastructure offering a consumer neobank app and a B2B payments API called Rafiki.

About this role

👋 About Us NALA is building Payments for the Next Billion. Faster, smarter, and fairer transfers for everyone. Since 2022, we've grown our business 120x, grown the team from 9 to 150+, raised $50M+ from top-tier investors , and were named to the Forbes Fintech 50 in 2025 & 2026. We operate two core products: NALA , our consumer app makes cross-border payments cheaper, faster and more reliable for the global diaspora. Allowing users to send money from the UK, US and EU to Africa and Asia. Rafiki , our B2B payments infrastructure, is powering global payments for global giants like MoneyGram & Western Union. Our team includes alumni from Wise, Stripe, Monzo, Revolut, and CashApp, operators who've scaled world-class products. We act with urgency, think deeply, and put our customers first always. At NALA, this isn't just a job. It's ownership, impact, and the chance to change global payments forever. Join us in building Payments for the Next Billion . 🙌 Your Mission We are rebuilding NALA's data foundation for a world where the thing asking the question might not be a person. Analysts, operators and executives already depend on our warehouse. Increasingly, so do agents. And an agent cannot ask a colleague what a column means, cannot tell that a join has silently doubled a number, and cannot sense that a metric is defined two different ways in two different places. Data that is merely good enough for a careful human is not good enough for that. The gap between the two is where this role lives. As Senior Analytics Engineer, you will own a set of domains in our transformation layer (dbt and Snowflake) and make them genuinely dependable, well modelled, tested, documented and semantically rich enough that both a person and a machine can get a trustworthy answer without a human in the middle to interpret. You will work alongside our existing Analytics Engineer, and between you set the modelling standard the rest of the team builds against. This is a role for someone who has used dbt to solve real business problems, not just to build what a ticket specified, and who has a considered view on what dimensional modelling and the modern warehouse actually demand now that AI is doing the querying. 🎯 Your Responsibilities in this Role Own the modelling and transformation of your domains end to end in dbt and Snowflake, from the business question through to the governed model that answers it Set and defend modelling methodology alongside our existing Analytics Engineer, grain, dimensional patterns, and where the exceptions genuinely earn their place Make your models agent-ready: semantically rich, consistently named, described in business terms, and safe for something to query without a person checking the output Own streaming and near-real-time pipelines (Kafka or similar) alongside batch transformation, keeping real-time flows reliable and cost-efficient before they become a bottleneck Establish and enforce coding and agentic-coding standards, systematic testing and documentation as CI-enforced defaults Optimise warehouse performance and cost, finding and fixing the query patterns and materialisation choices driving unnecessary spend Deploy and supervise autonomous agents against our data stack, and hold our data infrastructure to proper engineering standards: monitored, wired to incident.io and PagerDuty, with escalations, SLAs and documented processes Support and mentor analysts on analytics-engineering practice, raising the engineering standard across the team 🔥 Must-have requirements You solve business problems with dbt. You can walk us through modelling work you owned that changed a decision — what the business was getting wrong, how you structured the answer, and what you deliberately chose not to build. Three or more years of production dbt is the floor; what we are actually assessing is judgement. You have a real position on modelling methodology. Grain, facts and dimensions, fan-out and double counting are working tools rather than vocabulary. You can argue dimensional modelling against one big table in both directions and say where each belongs. You understand what the warehouse has to become in the AI age. You have a practical view of what makes a model trustworthy for an agent to query, and you have thought about it because you have tried it, not because you read about it. AI-native, beyond your editor. You use Cursor, Windsurf or Claude Code daily and you have built, deployed and maintained something agentic that runs without you watching it. You experiment outside work with tools your employer doesn't use, because you want to know how they differ. This is the requirement we weight most heavily. Streaming ownership. You have built and owned streaming or near-real-time pipeline infrastructure (Kafka or similar), and can talk about what broke as readily as what worked. Strong SQL, Python and warehouse fundamentals , with Snowflake or Databricks experience including query performance tuning and cost optimisation. Track record of engineering standards in dbt — testing, CI/CD, documentation and PR review workflows that outlasted you. Comfortable owning a roadmap. You can assess the current state, propose a plan and execute without being directed step by step, and you scope before you dive in. 💪 Nice to have requirements A data-engineering or software-engineering background before moving into dbt — it adds rigour and variety to a team drawn mostly from the analytics side Semantic-layer experience (Cube, Snowflake Semantic Views, dbt Semantic Layer) and an understanding of how governed metric definitions sit on a transformation layer Experience with Hex or a similar modern BI and notebook platform Fintech, payments or another regulated environment where data accuracy carries real consequences Familiarity with experimentation frameworks and product analytics ✅ Success in the role looks like 3-Month Metrics You own your domains outright and know where the bodies are buried in them — the models pe