AI Infrastructure Engineer

Pivot Search; Singapore; On-site; SGD 120,000 - 180,000; Full time

Job summary

Pivot Search is recruiting for a role building infrastructure to support long-horizon AI agents and an agent-driven distribution platform. The engineer will develop Python-based runtime and model routing, focusing on reliability, observability and cost effectiveness as the system scales from a working prototype to production.

Join a small founding-stage team shaping core product foundations, with opportunities to influence both architecture and deployment strategies while tackling real-world

Qualifications

• Proven ability to write production-grade Python code.
• Experience with distributed systems and scalable infra.
• Experience owning systems through real incidents and failures.
• Strong judgement on reliability, observability, latency and cost.
• Hands-on capability to shape architecture and deploy.

Responsibilities

• Build agent runtime and model-routing infra in Python with persistence and recovery.
• Design event and trajectory systems linking conversations, actions and outcomes.
• Own deployment foundations, observability, latency and cost.
• Challenge architectural assumptions and raise engineering velocity of a small team.

Skills

Python

Distributed systems

Production infrastructure

Reliability & observability

Architecture & hands-on coding

Tools

Kubernetes

Docker

Job description

Our client is building long-horizon AI agents and an agent-driven distribution platform, starting with relationship-native e-commerce. Its goal is to help people make better decisions by learning individual preferences, context and outcomes over time - then connecting genuine needs with the right products and services. The company recently completed a seed round of nearly US$10 million and is preparing for launch in the United States, with North America as its first priority before Southeast Asia and South America. This is a small founding-stage team where new hires can shape the core product and technical foundations.

About the Role

An AI assistant that supports people over weeks or months must remember context, execute tools reliably, recover from interruptions and connect later outcomes to earlier decisions. Our client is looking for an infrastructure engineer to make that experience dependable, observable and affordable at scale.

Responsibilities

• Build agent runtime and model-routing infrastructure in Python, including persistent state, reliable tool execution and recovery.
• Design event and trajectory systems that connect conversations, actions and outcomes, giving researchers usable learning data and engineers clear failure visibility.
• Own foundations for deployment, observability, latency and cost as the product moves from a working system to dependable production infrastructure.
• Challenge architectural assumptions, make practical trade-offs and raise the engineering velocity of a small team.

Required Skills

• Strong software engineering skills in Python.
• Deep experience with distributed systems and production infrastructure.
• Evidence of owning systems through real failures, operational incidents and scaling trade-offs.
• Strong judgement across reliability, observability, latency, cost and developer experience.
• Hands‑on ability to write production code and shape architecture.

Preferred Skills

• Comfort working in a small, fast-moving startup where the problem and solution evolve together.
• Interest in consumer AI, e-commerce, personalisation and long-horizon agent behaviour.
• Experience with LLM or agent runtimes, model routing, workflow engines, event systems or ML infrastructure.
• A project or system you can discuss in depth, including the decisions you would make differently today.