Gabriel
Schull
Building products from 0 to scale — across AI,
infrastructure, and full-stack systems.
About
I'm a fullstack engineer who thrives at the intersection of AI and product — building systems that scale from day one to millions of users.
At Stan, I'm the core engineer behind Stanley for Instagram, an AI product that went from concept to $2M+ ARR. Previously at Foreplay.co, where I was an early engineer scaling a 40M+ ad intelligence platform.
I care about shipping fast, building robust infrastructure, and creating products people actually love using.
Experience
Senior Fullstack Engineer
Core engineer on a 0-to-1 AI product (Stanley for Instagram), architecting and shipping the platform end-to-end — scaling from launch to $2M+ ARR.
- Designed AI agent infrastructure — streaming delivery, multi-agent routing, and extensible tool-calling framework
- Built content discovery pipeline with outlier detection, LLM analysis, and configurable relevancy scoring
- Built a multimodal content moderation system using LLMs with video frame extraction and text analysis (captions + transcripts) to filter unsafe content, alongside an LLM-inferred language detection and filtering pipeline for internationalization
- Led Instagram Graph API migration with full OAuth flow and graceful API fallback system
- Implemented data infrastructure with Airflow DAGs, S3 management, and performance observability
Fullstack Engineer
Early engineer on an ad intelligence platform with a 40M+ ad library, shipping across the full stack.
- Reduced cloud function costs by 60% via Kafka pipeline to dedicated Hetzner servers
- Architected Instagram ad capture using containerized Selenium with Chrome debugger API
- Built data ingestion pipeline — automated brand tracking, AI processing, ClickHouse aggregations
- Scaled infrastructure for 100%+ YoY user growth and 5M+ ads processed monthly
- Built publicly accessible library with programmatic SEO for increased web traffic
Previously: Research Analyst at Université du Québec (INRPME) · Analyst at CIBC · Campaign Coordinator at Rouge Media Group
Projects
Keith
Background coding agent platform that spawns sandboxed AI environments to write code, create PRs, and report results — triggered from Slack, GitHub, or a web UI.
Facial Verifier
Siamese neural network for one-shot facial recognition, inspired by the Koch, Zemel & Salakhutdinov architecture.
Transformer Translation
Custom transformer architecture for English-to-Italian translation, built from the 'Attention is All You Need' paper.
