AI Ad Performance Engineer: Video Insight & ROI

AI Case Leads; Pakistan; Remote; PKR 1,800,000 - 3,000,000; Full time

Job summary

AI Case Leads is seeking a skilled ML/software engineer to advance an AI platform that optimizes video ads for signed cases. You will connect Meta, Google and TikTok data with CRM results, building models that explain why ads perform and where to improve.

You’ll develop pipelines, multimodal models and explainable outputs, delivering concrete editing guidance and budget recommendations in a fast-paced, results-driven environment.

Qualifications

• Hands-on paid ads experience using Meta Ads or Google Ads data.
• Built or run systems judging ads/campaigns on calls, leads or sales.
• Experience with video or audio ML in production.
• Ability to explain outcomes with small data and avoid bias.
• Evidence of test sets and measured model accuracy.
• Experience with multimodal LLMs and cost/latency trade-offs.
• Strong Python and software engineering, data pipelines and Postgres.

Responsibilities

• Judge ads by actual revenue impact: calls, qualified leads and signed cases.
• Provide early warnings on fatigued ads and budget shifts across platforms.
• Develop video understanding at scale with shot/scene features.
• Link features to watch time and signed cases with confidence estimates.
• Offer editing guidance tied to exact seconds and scenes.
• Design briefs from evidence and generate variant scripts.
• Create evaluation sets and agreement checks with internal labels.

Skills

Paid ads data

Python

Postgres

TypeScript

ML engineering

Data pipelines

Next.js

Tools

Postgres

TypeScript

Next.js

Job description

AI Case Leads is seeking a skilled ML/software engineer to advance an AI platform that optimizes video ads for signed cases. You will connect Meta, Google and TikTok data with CRM results, building models that explain why ads perform and where to improve.

You’ll develop pipelines, multimodal models and explainable outputs, delivering concrete editing guidance and budget recommendations in a fast-paced, results-driven environment.