Senior engineering on Fox Atlas
I joined Fox Corporation as a Senior Software Engineer on Atlas, a media contextualization platform built to understand Fox content at scale and make that context useful for advertising workflows. The product sat at the intersection of media ingestion, ML analysis, cloud infrastructure, and the internal tools needed to turn derived context into usable segments.
The work mattered because the value of the platform depended on a continuous loop: new content had to enter the system, analysis had to produce useful metadata, and the resulting data had to be available quickly enough for teams building ad segments.
Fox later demoed Atlas publicly here: https://www.youtube.com/watch?v=TLyp2O98nQs
Event-driven media processing
The technical center of the work was an event-driven pipeline running across a suite of AWS services. Media content was continuously ingested, analyzed with ML, and made available downstream for product and advertising use cases.
I worked across Go services, TypeScript, React, and Terraform, which meant the implementation work crossed runtime code, user-facing surfaces, and infrastructure. That was important for this kind of platform: reliability was not only a backend concern, and the product workflow was not only a frontend concern. The useful architecture had to connect ingestion, analysis, data access, and deployment in a way the team could keep moving with.
From ML output to usable segments
The platform's goal was not simply to run analysis over content. The analysis needed to become structured context that could support segment building and help serve more relevant ads. That made the boundaries between content processing, metadata modeling, and product access especially important.
My approach was to keep the system oriented around the data's lifecycle. Events represented movement through the pipeline, services handled focused responsibilities, and infrastructure definitions kept the cloud footprint repeatable. The frontend and TypeScript work then helped expose the resulting data through product surfaces rather than leaving the value trapped in backend processing.
Shipping inside a cloud platform
Terraform and AWS were a core part of the delivery posture. For a platform built from managed services, infrastructure choices shape the application as much as code does. I treated DevOps as part of the product work: the pipeline needed to be deployable, observable, and understandable by the engineers operating it.
The engagement was hands-on engineering across a short, high-context window from April 2022 to December 2022. The practical outcome was progress on Atlas as a media intelligence platform: content moved through continuous ingestion, ML generated contextual signals, and those signals became available for advertising segment workflows.