Confirmed on the employer's own hiring board on Nov 16, 2026. First seen by Alion on Nov 13, 2026.
Principal Claude Code Engineer
Employment Type: Full-Time
Location: Lahore, Pakistan (Remote)
About the Role
Pixalate is hiring four Principal Claude Code Engineers in Lahore, Pakistan. Our engineers already write code with Claude Code every day. This role is for the engineer who does it best: you ship production code mostly through Claude Code, you set up the context that makes it accurate for each codebase you work on, and you are accountable for what reaches production. The bar is the quality of the code Claude Code produces under your direction and the engineering judgment you apply to it. It is not a particular programming language, and it is not a research degree. This is not a model-training, data-science or prompt-tinkering role.
You will work across all of Pixalate's products as priorities shift, from the data pipelines behind our Media Ratings Terminal (MRT) to our fraud and invalid-traffic (IVT) detection and the MCP servers our clients connect to.
You will work with engineers and product leads in the US, Singapore and Pakistan. Our Senior AI Platform/System Engineer, a separate role, is building shared coding agents hosted in the cloud, so the agent setup you create can be shared with your teammates. You can start without it.
Key Responsibilities- Ship production code through Claude Code every day: plan the change, direct the agent, and own the result through review, testing and release.
- Write and maintain the CLAUDE.md files, skills, subagents and MCP configuration that give Claude Code each product's code, business rules and workflows, and keep them current as the code changes.
- Review agent-written code for correctness, security, cost and maintainability, and write the tests and checks that prove it works before it ships, such as comparing output against the current version on a real day of data.
- Take work from ticket to production with the team: pull requests, code review, UAT and release, with a written plan when a change affects client-facing data.
- Work on whichever product needs you: data pipelines and BigQuery jobs, the MCP servers and APIs clients use, and fraud and invalid-traffic detection.
- Keep what you ship accurate, reliable and affordable: validate output on real data, measure detection quality on real traffic, and keep BigQuery cost and runtime under control.

