As a Backend Software Engineer (IC3) in the AI & Data Products squad, you will design, build, and operate backend services that power AI-native and ML-native product features.
Your mission is to help Spendesk move from isolated intelligence components to real, user-facing product capabilities. In practice, this means partnering with ML Engineers to productionize predictive logic, expose it through clean APIs and services, and integrate it into workflows that automate tasks, simplify decision-making, or anticipate user needs.
You may work on features such as:
automated categorization and enrichment of spend-related workflows
Задачи
Backend services for AI and ML-powered product features
Design, build, and operate backend services and APIs that power AI-driven, ML-driven, or automation-heavy product capabilities
Translate predictive logic and AI outputs into reliable backend behaviors that can be consumed by user-facing product flows
Build the service layer that allows intelligent features to be integrated into real workflows with strong standards on latency, reliability, and security
Ensure features are designed for production, not just experimentation, with clear ownership of deployment, monitoring, and maintainability
Productionization of ML and LLM capabilities
Partner closely with the squad’s ML Engineers to productionize predictive models and LLM-driven capabilities
Integrate model-serving APIs or LLM calls into robust backend services (your squad, or the applicative squad’s services) with proper retries, fallbacks, and observability
Help define evaluation and monitoring patterns that make intelligent product behaviors measurable over time
Contribute to the engineering patterns that allow ML and AI capabilities to be reused across multiple product features
Automation, prediction & workflow simplification
Build backend capabilities that help automate repetitive tasks, anticipate user needs, or simplify complex workflows
Work on product experiences where AI or ML can reduce manual effort, improve decision quality, or shorten time to value for users
Partner with Product and Design to turn ambiguous ideas into concrete backend implementations with measurable impact
Bring pragmatism to delivery, balancing experimentation speed with long-term maintainability and trust