End Point Engineer

About PointFive

PointFive is the AI Efficiency OS. From the cloud to the coding agent, we're the only platform that manages AI spend everywhere it happens.

Engineering and FinOps teams use PointFive to make their organizations more efficient and their cloud and AI more effective. We don't just show what you spend. We show what you're wasting, and we fix it autonomously. NuBank saw ROI in 10 days. Customers average 1,200%+ ROI and a 4.9 rating on G2.

Founded by the team behind IntSights (acquired by Rapid7), PointFive recently closed a $60M Series B led by Accel, with participation from Entrée Capital and Salesforce Ventures.

About The Role

TokenShift is PointFive's newest product — a lightweight CLI proxy that sits between AI coding agents and the shell or server, routing to different models for perfect model to effort fit, compressing command output to cut token spend and surface fleet-wide visibility. As an Senior software Engineer (Endpoint), you'll own the layer that makes TokenShift work: the per-agent harnesses, the compression rules and strategies, and the installation and hook mechanics that get us onto developer machines — reliably, invisibly, and at scale. This is a rare chance to build foundational infrastructure for AI developer tooling, at a company where your work directly shows up in the product used by engineering teams at companies like Expedia and EON.

What You'll Do

• Own the installation and enrollment mechanics — the CLI flows, MDM scripts, and platform-specific packaging (Windows msi/exe, macOS pkg, Linux deb/rpm) that put TokenShift on developer machines at fleet scale.
• Instrument and monitor endpoint behavior through the telemetry pipeline, using signals to improve coverage and catch regressions.
• Build and maintain a robust lab enabling automatic research and development of new capabilities and platforms by AI agents.
• Build and maintain agent harnesses — the per-agent adapters that hook into Claude Code, Cursor, Copilot CLI, Codex, and future agents to route network and shell commands through the TokenShift proxy.
• Develop and extend compression rules and strategies for new command types (git, build tools, infra CLIs, search, and more), ensuring token reduction never comes at the cost of semantic accuracy.
• Maintain correctness and resilience of the proxy layer — compression must be reversible on demand, hooks must never block a developer's command, and failures must be silent and transparent.
• Collaborate with the research team to translate newly supported agent architectures into production-grade harnesses.
  
  

• Solid understanding of Windows and Unix fundamentals: process model, shell execution, stdin/stdout/stderr, file permissions, and IPC.
• Deep familiarity with MDM deployment patterns (Jamf, Intune, JumpCloud, Ansible).
• Strong proficiency in Go — the TokenShift CLI is written in Go, and this is where the core work happens.
• Experience building or maintaining CLI tools in production environments.
• Attention to correctness and edge cases — you think about what happens when things go wrong before they do.
• Ability to work across platforms (macOS and Linux) and understand their differences at the system level.
  
  

• TypeScript experience.
• Hands-on experience with AI coding agents (Cursor, Claude Code, Copilot CLI, Codex) as a user or developer.
• Experience with semantic compression, summarization, or output parsing.
  
  

• Founders with a track record
  
  

• Category-defining product
  
  

• Early-stage leverage
  
  

• You'll build the infrastructure layer for AI developer tooling