Senior Staff Engineer - Agent Platform

📋 Description

• • As a Senior Staff Software Engineer at tem, you will play a pivotal role in shaping and scaling our cutting-edge Agentic layer, a core component of our AI-native transaction infrastructure. This position is designed for an experienced engineer who thrives on both deep, hands-on technical execution and strategic, organizational leadership.
• • Your primary focus will be on building the foundational elements that enable AI agents to operate reliably and efficiently in a production environment. This includes developing the core runtime and associated tooling, establishing robust integration interfaces with our existing systems, and defining the engineering standards that empower teams to deploy agentic capabilities safely, reliably, and repeatedly.
• • You will be involved in the entire lifecycle of agentic systems, from initial pilot phases through to full-scale production roll-outs. This requires close collaboration with cross-functional teams, including engineering, product, data science, and domain experts, to translate complex real-world workflows into durable, AI-powered solutions.
• • A key aspect of this role is your ability to influence technical direction across teams, ensuring that architectural designs are tightly aligned with product roadmaps and overarching business objectives. You will need to exercise sound judgment to strike a balance between the need for rapid delivery and the imperative of maintaining long-term system integrity and scalability.
• • You will be responsible for shipping flagship agentic capabilities, delivering high-impact workflows from conception to production with clearly defined success metrics and agile iteration cycles. This involves taking ownership of the end-to-end development and deployment process.
• • A significant part of your contribution will be in building and operating production-grade agent systems. This entails designing systems that are not only reliable but also predictable under real-world constraints such as latency, cost, data quality variations, and potential failure modes. You will implement strong patterns for state management, idempotency, and safe recovery mechanisms.
• • You will create and champion shared foundational elements that enable other teams to build agents consistently. This includes developing core primitives like runtime patterns, standardized tool interfaces, effective context management strategies, and reusable shared libraries, thereby avoiding fragmented, one-off solutions.
• • Establishing a pragmatic Agent Development Life Cycle (ADLC) will be crucial. You will implement robust evaluation frameworks, safety guardrails, comprehensive tracing capabilities, effective monitoring systems, and streamlined release processes to ensure agents can be continuously measured, debugged, and improved.
• • You will integrate Machine Learning (ML) and Large Language Model (LLM) components into production workflows. This involves close partnership with ML and Data teams to productionize models and LLM capabilities, ensuring clear contracts, versioning, observability, and safe degradation patterns are in place.
• • As agent adoption scales, you will maintain clear domain boundaries, defining shared semantics for agent tools and data access. This prevents domain drift and ensures consistency while empowering teams to innovate rapidly.
• • You will collaborate closely with the Platform engineering team to adopt and extend existing infrastructure and developer tooling, including CI/CD pipelines, developer experience (DevEx) enhancements, and observability systems. Your work will contribute back to these systems where agentic workloads introduce novel requirements.
• • Success in this role will be measured by tangible outcomes: within 3 months, shipping the first flagship agentic workflow to production with defined KPIs, runbook/on-call ownership, and baseline telemetry; within 6 months, shipping additional workflows and implementing a lightweight ADLC including evaluations, guardrails, monitored rollouts, and rollback capabilities; and within 12 months, demonstrating repeatable agent development by enabling 2+ product teams to ship on shared foundations, accelerating time-to-production for new agents, and consistently meeting reliability and cost targets.