AI-Powered Ad Scheduler

Publicis Sapient — Software Engineer Intern · May 2025 – Aug 2025 · Chicago, IL

2,000+ locations ~85% faster (20 → 3 min/ad) 96.8% test coverage

Overview

I developed a full-stack, AI-integrated scheduling platform to optimize advertising across 2,000+ retail locations, cutting setup time from ~20 minutes to ~3 minutes per ad — an ~85% efficiency gain.

What I Built

Geolocation services to map and manage 2,000+ store locations for campaign targeting.
LLM-based features for tailored scheduling suggestions and compliance reminders.
Robust validation and error handling for media files and time slots.
Campaign performance analytics for post-launch insights.

Outcomes

~85% faster scheduling (≈20 → ≈3 minutes per ad).
96.8% automated test coverage for reliability and smoother deployments.
Cloud-hosted architecture for scalable, secure remote access.

Responsibilities

Engineered full-stack features and data models for campaign scheduling.
Designed and optimized PostgreSQL schemas for metadata and asset management.
Integrated LLM prompts/workflows to improve scheduling accuracy and compliance.
Implemented automated testing to ensure high coverage and system reliability.

Back