How to apply for Backend / Platform Engineer, AI Analytic Engines

About Thirdlaw

Thirdlaw builds infrastructure for enterprises to monitor, evaluate, and control AI behavior at runtime. The company focuses on detecting safety, compliance, and operational risks in LLMs and AI agents that traditional observability tools miss. This is a chance to work on the control layer for enterprise AI, backed by top-tier venture firms and enterprise design partners.

About the role

This role involves architecting scalable, low-latency services for running AI evaluations in real-time and batch. You will integrate with streaming data pipelines and trace-based event models. The work directly shapes how enterprises monitor and control AI behavior where it matters: at runtime.

A typical day

A typical day might involve designing or debugging low-latency services that run AI evaluations in real-time and batch. You could work on integrating streaming data pipelines or trace-based event models. You might also collaborate with enterprise design partners to ensure the system detects and responds to unsafe AI behavior at runtime.

Who Thirdlaw is looking for

• Experience building low-latency backend services that handle streaming data and event-driven architectures.
• Familiarity with distributed pipelines and reconciling trade-offs between latency, correctness, and observability.
• Interest in AI safety, evaluation, and runtime control of LLMs or AI agents.
• Comfort working in a remote, full-time engineering role with enterprise design partners.

Tips for this application

• Highlight any experience with real-time or batch evaluation systems, especially those involving AI or LLMs.
• Mention specific projects where you integrated streaming data pipelines or trace-based event models.
• Show you understand the difference between traditional observability (latency, cost) and behavioral AI monitoring (bad decisions, data leaks).
• Reference Thirdlaw's focus on runtime control and the question 'Is this OK?' in your materials.
• If you have worked on systems that detect or stop unpredictable AI behavior, describe that work in detail.

What to cover in your cover letter

['Your experience building low-latency services that process streaming or event-based data.', 'How you have handled trade-offs between latency, correctness, and observability in distributed pipelines.', 'Why you want to work on AI evaluation and control at runtime, not just model tuning or log piping.', 'Any direct experience with AI agents, LLMs, or enterprise safety/compliance systems.']

Research before applying

• Read Thirdlaw's website and any public materials on their control layer for enterprise AI.
• Understand the difference between traditional observability and AI behavior monitoring (safety, compliance, unpredictable actions).
• Look into the company's venture backing and enterprise design partners to understand their stage and market.
• Research common failure modes of LLMs and AI agents in enterprise settings, such as data leaks or bad decisions.

Likely interview topics

Based on the job description, expect questions about:

• How to architect a low-latency service for real-time AI evaluation that also supports batch processing.
• Designing trace-based event models for monitoring AI agent behavior.
• Strategies for reconciling latency, correctness, and observability in distributed pipelines.
• How to detect when an AI model makes a bad decision or leaks sensitive data at runtime.
• Experience integrating streaming data pipelines with evaluation engines.

Common mistakes to avoid

• Focusing only on generic backend skills without connecting them to AI evaluation or runtime control.
• Treating the role as a standard observability or logging position rather than AI behavior monitoring.
• Ignoring the enterprise context and the need for real-time action when AI behavior is not OK.