Pixalate seeks an AI Prompt Engineer to build and maintain production AI agents that investigate invalid traffic and provide analytics. The role involves authoring and maintaining agent skills, system prompts, and plugins, designing agents that interactively gather requirements, versioning changes, debugging agent failures, defining evaluation criteria, building ground-truth datasets, and ensuring safety and trust in agent behavior. Candidates must have experience shipping LLM agents, proficiency in Python or TypeScript, strong written English, understanding of tool-calling and prompt injection risks, and preferably exposure to ad-tech or fraud detection. The position is full-time, remote in Pakistan, with some US timezone overlap required. Benefits include competitive compensation, internet reimbursement, flexible hours, and professional development opportunities.
What you'll do
Author and maintain Claude skills and plugins defining agent capabilities and behaviour
Design agents that ask clarifying questions and gather requirements interactively
Similar jobs
More roles worth a look
Related opportunities based on specialty and working model so candidates can keep momentum.
Decide what belongs in a skill, system prompt, or tool
Maintain the skill library as products evolve
Own system prompts behind client-facing agents (IVT investigator agents, IQ Agent / Analytics MCP, MRT MCP)
Version prompts and skills with clear records of changes
Debug agent failures such as wrong tool selection, hallucinated findings, missed clarifying questions, context exhaustion
Define correctness criteria for agents and build evaluation suites
Build ground-truth datasets from real customer scenarios and measure changes
Catch regressions before customers do
Engineer context strategies for agent input order and budget
Collaborate with engineering on MCP tool surfaces and tool descriptions
Tune for cost, latency, and accuracy
Guard against prompt injection, data leakage, and unsafe tool use
Ensure agents represent findings accurately without overstating confidence
Requirements
Demonstrable experience building LLM agents shipped to real users
Hands-on work with agent skills, plugins, system prompts, or equivalent agent-behaviour tooling
Experience building evaluation suites and ground-truth datasets
Strong written English
Working proficiency in Python or TypeScript
Comfort reading the codebase agents run in
Understanding of tool-calling / function-calling and agent tool selection
Awareness of prompt injection, PII handling, and failure modes of customer-facing AI
Comfortable with some U.S. timezone overlap for team meetings
Ad-tech, digital advertising, or fraud-detection exposure preferred (programmatic, DSPs, SSPs, CTV, IVT, ad fraud concepts) or willingness to learn deeply in first 90 days
Experience with MCP servers or similar tool-integration standards is a plus
Experience with observability and tracing for LLM systems is a plus
Experience with retrieval systems and their failure modes is a plus
Data analysis background including SQL and working with large datasets is a plus
Competitive Compensation PackageMonthly Internet ReimbursementCasual, Remote Work EnvironmentFlexible HoursOpportunity for Advancement and Professional DevelopmentFun Annual Team EventsBeing part of a high-performing global team
Apply now
Ready to take the next step in your career? Click the button below to continue to the application process.
