P

Pixalate

SaaS platform for ad fraud analytics

AI Prompt Engineer - Skills & Prompts

AI Application EngineeringConfirmed live in the last 24 hours
Full-timeRemote in Pakistan

About the Job

Requirements

  • A Bachelor's degree in Computer Science, AI, Machine Learning, or a related field, or an exceptional research track record.
  • Demonstrable experience building large language model agents that have shipped to real users, rather than prototypes or demos.
  • Hands-on experience with agent skills, plugins, system prompts, or equivalent agent-behaviour tooling.
  • Ability to describe an evaluation suite built, its ground truth, and a regression it caught.
  • Strong written English.
  • Working proficiency in Python or TypeScript and comfort reading the codebase the agents run in.
  • Understanding of tool-calling or function-calling and how agents select tools.
  • Awareness of prompt injection, personally identifiable information handling, and failure modes of customer-facing artificial intelligence.
  • Comfort with some U.S. timezone overlap for team meetings.

Responsibilities

  • Author and maintain Claude skills and plugins that define agent capabilities and behaviour.
  • Design agents that ask clarifying questions and gather requirements interactively rather than responding from incomplete information in a single turn.
  • Decide what belongs in a skill, a system prompt, or a tool.
  • Maintain the skill library as products evolve.
  • Own the system prompts behind client-facing agents, including IVT investigator agents, IQ Agent / Analytics MCP, and MRT MCP.
  • Version prompts and skills and maintain a clear record of what changed and why.
  • Debug agent failures including wrong tool selection, hallucinated findings, missed clarifying questions, and context exhaustion.
  • Define correctness criteria for each agent and build evaluation suites to test them.
  • Build ground-truth datasets from real customer scenarios and measure changes against them.
  • Catch regressions before customers do.
  • Engineer context strategies, deciding what the agent sees, in what order, and within what budget.
  • Work with engineering on MCP tool surfaces, including which tools exist, how they are described, and how descriptions support correct tool selection.
  • Tune agents for cost and latency as well as accuracy.
  • Guard against prompt injection, data leakage, and unsafe tool use in customer-facing agents.
  • Ensure agents represent company findings accurately and do not overstate confidence.

Desired Qualifications

  • Exposure to ad technology, digital advertising, or fraud detection, including programmatic, demand-side platforms, supply-side platforms, connected TV, invalid traffic, or ad fraud concepts; candidates without this experience should be prepared to learn the domain deeply in their first 90 days.
  • Experience with Model Context Protocol servers or similar tool-integration standards.
  • Experience with observability and tracing for large language model systems.
  • Experience with retrieval systems and their failure modes.
  • A data analysis background, including SQL and working with large datasets.

About the Company

Pixalate

SaaS platform for ad fraud analytics

Pixalate provides fraud protection, privacy, and compliance analytics for the digital advertising industry. Its SaaS platform analyzes ad traffic across Connected TV, mobile apps, and websites to detect invalid traffic (IVT) in real time, with blocklists and reporting to help optimize campaigns. It differentiates itself through cross-channel coverage, MRC accreditation for SIVT detection, and detailed traffic-quality insights. Its goal is to help advertisers and publishers protect ad spend, uphold brand safety, and ensure transparent ad delivery by preventing fraud.

Company size
51-200
Company stage
Growth Equity (Venture Capital)
Total funding
$27.5M
Headquarters
McLean, VA
Founded
2012

Recently Posted Jobs

Explore more opportunities at Pixalate

See more jobs