A Chicago-area media archive and licensing company came to Armakuni for gen AI POC for automated metadata tagging and clip discovery from long-form archival video; storage modernization assessment for petabyte-scale migration from TrueNAS and Azure Blob to AWS S3 Intelligent-Tiering.
The challenge
Armakuni went from an intro call with Randy Burgess, VP of Technology, to MAP Lite funding approved and a Gen AI POC in motion in under 2 months. Two SOWs were delivered within 48 hours of the first meeting. The approach combined a Gen AI POC demonstrating tagging automation with a storage modernization assessment and clear execution roadmap, aligning with the CIO's directive to consolidate vendor dependencies. Armakuni was selected as the partner for both innovation and execution.
Scale: Manages approximately 1.2 PB of media across on-premises TrueNAS (~900 TB) and Azure Blob (~300 TB); stock footage archive licensed through Getty Images taxonomy.
What we built
Armakuni delivered gen AI POC for automated metadata tagging and clip discovery from long-form archival video; storage modernization assessment for petabyte-scale migration from TrueNAS and Azure Blob to AWS S3 Intelligent-Tiering. The engagement ran as two fixed-scope SOWs: (1) MPI Media Assessment: storage infrastructure audit, AWS architecture blueprint, MAP-aligned migration readiness report, and Phase 2 execution SOW; (2) SFA Automation POC: Gen AI clip identification and metadata tagging proof of concept. AWS funded the work at $5,700.
The outcomes
5 measurable outcomes shipped across the engagement. The ones that moved the business the most:
10 to 20 times more content processed for tagging per unit of editorial time, compared to the fully manual workflow. The Gen AI pipeline processes a video, identifies candidate clips, generates descriptions, validates tags against the Getty taxonomy, and surfaces only the edge cases requiring human judgment. The editorial team reviews AI output rather than producing it from scratch…
Subtitle generation from Indian-language archival content to English demonstrated in the POC, opening a category of content that was previously undiscoverable. because it existed only in non-English audio with no transcription.
MAP Lite funding secured within 8 days of first contact, covering the full cost of both the Gen AI POC and the storage assessment.. MPI Media's total cost to the customer was $0 across both SOWs.
Storage migration readiness plan delivered in 2 to 3 weeks, covering the full 1.2 PB archive across TrueNAS and Azure Blob with a phased execution. roadmap, security gap analysis, and TCO comparison that gave leadership the information needed to commit to the migration phase.
Amazon S3 Intelligent-Tiering architecture designed to eliminate manual lifecycle decisions permanently.. Once the archive moves to S3 Intelligent-Tiering, storage cost automatically tracks actual usage. No one needs to decide which assets belong in which tier. The platform adapts as access patterns change, with no retrieval delay and no ongoing manual audit required.
I wanted to send a huge thank you for your tremendous partnership with MPI Media Group. We went from an intro call with their VP of Technology, having never spoken to the customer before, to MAP Lite funding approved and a Gen AI POC in motion, all in under two months. It was a truly bar-raising partnership. Can't wait to continue working together on MPI Media Group and many more customers to come\\!
Brittany Bonar, AWS Account Manager
Built on AWS
The production environment runs on 6 first-party AWS services. Delivered under our AWS Migration and Modernization Consulting, Generative AI Services Competency competencies.
What's next
Phase 2 migration execution: full 1.2 PB migration from TrueNAS and Azure Blob to Amazon S3 Intelligent-Tiering. Production deployment of the Gen AI tagging and clip discovery pipeline at scale across the full archive.