Kenyon Johnston

Welcome. I am a platform-focused Product Manager based in Atlanta, GA, specializing in AI, Cloud, SaaS, and MLOps. My career is dedicated to building the "intelligent engines" behind global platforms that reach millions of users.

I thrive on one question: How can we use predictive technology to revolutionize the content supply chain? I don't just ship features; I scale invisible, intelligent infrastructure that turns complex data into commercial opportunity.

With a decade of experience across Warner Bros. Discovery, IBM, and Acquia, I have navigated complex technical ecosystems. I leverage my background as a Linux Systems Engineer to speak the language of engineering teams while translating technical complexity into clear ROI for executive stakeholders.

The Vision

I build decision-intelligence environments. Whether it’s reducing operational overhead through automated metadata extraction or optimizing greenlight evaluations with predictive analytics, my goal is to ensure the underlying technology is invisible, intelligent, and invincible.

Currently pursuing an M.S. in AI and Machine Learning (2026), I apply academic rigor to solve real-world fragmentation in the media landscape through the development of studio-grade decision intelligence tools.

Verdict - Greenlight Intelligence Platform

The Challenge: Subjective, slow-moving greenlight evaluations in the film development cycle.

The Solution: A Predictive Intelligence Platform built with React, AWS, and LLM orchestration that quantifies creative potential and mitigates production risk.

• Explore Case Study

Release 1.0 Core Capabilities:

• Predictive Success Modeling: Utilizes a "Top Sheet" prediction engine to forecast project viability based on production budget, release strategy, and historical market data.
• AI Script & Treatment Audits: Automated diagnostic layers that evaluate narrative pacing, dialogue authenticity, and structural soundness before the first frame is shot.
• Executive Verdict Engine: A cloud-native submission portal that converts raw creative assets (PDF/DOCX) into quantitative marketability scores.

AirDate - Metadata Orchestration Platform

The Challenge: High-entropy streaming data and name ambiguity across fragmented global premiere schedules (e.g., 2006 vs 2026 series title collisions).

The Solution: A serverless RAG platform that synchronizes static metadata with real-time industry trade signals to provide high-fidelity "Briefing" intelligence.

• Launch Platform

Platform Engineering Highlights:

• Multi-Source RAG Architecture: Engineered a parallel execution engine (AWS Lambda/Python) that orchestrates retrieval across TMDB, NewsAPI, and Tavily to bypass training-data latency.
• ID-Aware Data Disambiguation: Implemented a strict ID-locking system between React and API Gateway to resolve title collisions and prevent AI hallucinations for multi-version franchises.
• Real-Time Sentiment Analysis: Leverages Amazon Bedrock (Claude 3.5 Sonnet) to quantify critical buzz from Rotten Tomatoes and Collider into a "Critics Choice" performance metric.
• Serverless Edge Deployment: Built for global scale using S3/CloudFront for static hosting with a custom Skeleton Loading layer to optimize perceived latency for RAG-heavy briefings.

I am always open to discussing the future of AI in Hollywood, multi-cloud architecture, or high-level product collaboration.

• Location

  Atlanta, GA

• Credentials

  CSPO, AWS Solutions Architect, AWS AI Practitioner, CompTIA Security+

• Email

  kenyon.johnston@gmoonmedia.com

• Elsewhere

  • LinkedIn
  • GitHub