Moss & Bone
Media & Entertainment
Prismcast
A streaming home for film that deserves better than an algorithm’s afterthought.
- Client
- Northern Reel Media
- Year
- 2020
- Duration
- 9 months
- Team size
- 7 engineers
Interactive Demo
A hands-on mock of the product we built — click around.
app.prismcast.com
Demo interface shown in English.
01
The challenge
Northern Reel’s 800-film catalogue earned pennies per view on advertising platforms, and its YouTube revenue had fallen 60% in two years as the algorithm moved on. The rights holders who licensed films to the platform demanded studio-grade DRM — Widevine and FairPlay at minimum — plus geo-windowing to respect territorial deals, none of which existed off the shelf at an indie budget. Launch day was fixed by a festival partnership and immovable, and the audience skewed towards cinephiles who would judge the service harshly on start-up time, subtitle quality and 4K bitrate. The company had six weeks of runway beyond launch and needed paid subscriptions to work from day one.
02
Our solution
We built the streaming pipeline on HLS and DASH with per-title encoding ladders tuned for film grain, served through a multi-region CDN with signed URLs, and integrated multi-DRM licence delivery covering Widevine, FairPlay and PlayReady. A Node.js API powers React web and mobile apps plus a TV client, with a geo-windowing engine that enforces each title’s territorial rights automatically. Stripe handles subscriptions, trials and gift passes, and a lightweight recommendation service blends editorial collections with collaborative filtering so catalogue depth surfaces without a data-science team. Playback starts in 1.2 seconds at the 95th percentile, and launch weekend served 4.1 million minutes with no manual scaling.
03
Results
120k
Subscribers in the first year
1.2 s
Playback start time at p95
4.1M
Minutes streamed on launch weekend
+58%
Month-3 retention vs target
Technology stack
HLS / DASHNode.jsReactAWS CloudFrontStripeFFmpegKubernetes
Facing a similar challenge?
We've solved this class of problem before. Let's talk about how that experience applies to your situation.
Discuss your projectNext case study
FurrowIQ →