No library migration, zero raw-video upload —
FAST channels built on your library, your library earning ad revenue.
A FAST (Free Ad-Supported Streaming TV) channel is a linear streaming channel viewers watch for free, earning from ad breaks. Point Media Tech's FAST (Kaster) is a hybrid-cloud MAM and FAST channel platform: your master files stay in your own facility and only 1–5% of derived data reaches the cloud — enough to search tens of thousands of hours by plain language, auto-generate captions with Whisper, play out locally, insert ads with SSAI, and output to FAST platforms. Your existing library becomes an ad-supported channel playing out 24/7.
0Raw files uploaded
1–5%Derived data to cloud
2 wkChannel to launch
1/40Cost vs. legacy MAM
* Derived data = audio summaries, keyframe thumbnails, and semantic embeddings. Upload share, launch timeline, and cost comparison depend on library scale, formats, and deployment environment.
The masters never leave the building.
The channel still monetises.
On Point Media Tech's FAST console, the schedule grid is a 24-hour channel built from the library — and the row that stops most launches is the rights window underneath it. The architecture panel on the right shows that only 2.2% of the library ever reaches the cloud.
From "migrate the whole library to the cloud"
to on-premise-first, cloud gets only the extracts.
Point Media Tech's FAST changes library monetization from "upload everything" to "on-premise first, cloud gets only the extracts". The old way: to put AI on your library and launch FAST channels, you first haul terabytes of masters into the cloud, pay steep storage and egress bills, and clear a data-residency security review — pushing a single channel two to three months out before it reaches a FAST platform. Point Media Tech FAST inverts that. An Edge Agent reads your existing NAS in place: masters are never uploaded, transcoded, or moved, and only 1–5% — semantic indexes, captions, and thumbnails — leaves for analysis. AI semantic search, Whisper captions, local playout, and SSAI ad insertion run on one platform, with output to FAST platforms within two weeks.
Bulk-migrate masters, burn bandwidth, wait on security review
Terabytes of masters must be uploaded before any AI can run; data-migration and egress costs inflate TCO; large-volume data egress triggers compliance review that slows onboarding; legacy perpetual-license MAM often starts in the millions.
On-premise-first, cloud extracts only, monetizing in two weeks
The Edge Agent reads your on-prem NAS with zero raw upload; only 1–5% of derived data leaves; VLM semantic search + Whisper captions; local playout keeps broadcasting through outages; SSAI / SCTE-35 ad insertion; one schedule outputs to multiple FAST platforms.
Today · Move the whole library to the cloud
- Terabytes of masters uploaded in bulk
- Heavy storage and bandwidth costs
- A security review before anything can leave
- Two to three months to get one channel live
After · On premise first, only derived data goes up
- An edge agent reads the NAS you already have
- No masters uploaded, transcoded or moved
- Only 1–5% — semantic index, subtitles, thumbnails — leaves
- Output to FAST platforms within two weeks
Two pillars, built on a hybrid cloud
where raw video never leaves your facility.
Point Media Tech's FAST rests on two pillars, AI media asset management (MAM) and FAST monetization, which share one on-premise semantic index, on a hybrid-cloud edge layer that keeps masters at zero upload. Each layer is sized independently, licensed by channel and module, and extensible as you grow.
AI media asset management · MAM
A vision-language model (VLM) builds a shot-level semantic index, so a plain-language phrase ("Taipei 101 at dusk") finds the frame across tens of thousands of hours — scored on five dimensions: people, object, scene, action, and emotion. Whisper AI auto-generates multilingual transcripts and SRT captions with built-in recursive correction; unnamed faces cluster automatically and match library-wide once named; the Edge Agent reads your existing NAS non-invasively, with no master migration or transcoding.
- Five-dimension VLM semantic search
- Whisper multilingual transcripts / captions
- Automatic face clustering & matching
- Non-invasive on-prem NAS integration
FAST TV monetization · Playout & Ads
A local playout engine runs scheduling and stream output independently of the cloud, so channels keep broadcasting even when the network drops. Define ad windows, intervals, durations, and pre/mid/post positions, and break rules auto-generate the EPG (JSON / XMLTV). SSAI server-side ad insertion stitches ads at the stream layer from SCTE-35 signaling, so client-side ad blockers have nothing to block and impressions can be reconciled for billing.
- Local playout · outage-proof broadcast
- Break rules auto-generate EPG
- SSAI / SCTE-35 ad insertion
- EPG JSON / XMLTV export
Hybrid-cloud edge & security · Hybrid Edge
The Edge Agent deploys on your existing NAS or a low-spec x86 host (Docker) — no GPU purchase, no new server room. Only audio summaries, keyframes, and semantic embeddings (about 1–5% of derived data) are encrypted over outbound-only HTTPS (443), so the corporate firewall opens no inbound ports. Uploads use presigned URLs valid for a single hour; cloud scratch data auto-deletes on a 30-day lifecycle. Not a single frame of the master leaves your facility.
- Zero raw upload · no library migration
- Outbound-only 443 · no inbound ports
- Presigned URLs · 1-hour expiry
- Cloud scratch auto-deletes in 30 days
FAST is a shortcut from library to viewer —
from a file landing on your NAS to earning ad revenue.
Point Media Tech's FAST channel build runs in five stages: ingest, AI cataloging, scheduling, ad insertion and distribution; a video moves from your on-prem NAS to multi-platform output with each step handed off by the system, and the master stays in your facility the whole time.
- IngestOn-prem NAS · fingerprint dedupe
- AI CatalogVLM index · Whisper captions
- SchedulePlaylist · break rules · EPG
- Ad InsertSSAI · SCTE-35 signaling
- DistributeSamsung / LG / Pluto / Roku
Engineering specifications.
Point Media Tech's FAST uses an on-premise-first hybrid-cloud architecture and outputs HLS streams and EPG JSON / XMLTV; a full spec sheet and live engineering walkthrough are available on request.
| Hybrid-cloud architecture | On-premise-first · Edge Agent on existing NAS / x86 (Docker) · zero raw-master upload · only 1–5% derived data egress |
|---|---|
| AI semantic search | Shot-level VLM semantic index · five-dimension scoring (people / object / scene / action / emotion) · natural-language search · automatic face clustering |
| Speech & captions | Whisper multilingual transcripts and SRT captions · recursive correction · one-click AI translation for global distribution |
| Scheduling & EPG | Playlist selection · break-rule auto-scheduling · EPG JSON / XMLTV export · platform-specific specs · multi-time-zone + loop |
| Ad insertion | Time-window break rules (start/end · interval · duration · position) · SCTE-35 signaling · server-side SSAI · bypasses ad blockers |
| Playout & output | Local playout runs independently · outage-proof broadcast · preview transcode H.264 / MP4 (faststart) · ad stream over HLS |
| Distribution platforms | Samsung TV Plus · LG Channels · Pluto TV · Roku Channel · one schedule, simultaneous multi-platform · own OTT |
| Security & governance | Outbound-only HTTPS 443 · presigned URLs (1-hour expiry) · cloud scratch auto-delete in 30 days · user / role / channel RBAC · full audit logs |
| Deployment & licensing | Cloud or on-premise · containerized · no GPU required · licensed by channel count and feature modules · monthly or annual contract |
Three steps, from evaluation to production.
Request a POC
Submit the contact form or request a 100-hour POC. Sales responds within one business day to schedule a technical consultation.
Security & integration assessment
Our consultant walks through the data flow and security architecture, and assesses on-prem NAS integration, AI semantic search, target platforms, and monetization strategy.
Deploy and go live
The Edge Agent installs on your existing NAS; after AI cataloging and platform onboarding pass acceptance, we phase your first channels live within two weeks, then scale.
Frequently asked questions.
What is a FAST channel, and how do broadcasters earn from free ad-supported streaming?
A FAST (Free Ad-Supported Streaming TV) channel is a linear streaming channel that viewers watch for free and that earns from ad breaks. It plays out 24 hours a day on a schedule, with breaks marked by SCTE-35 and ads stitched in by SSAI. Unlike subscription OTT, FAST needs no paid subscription system, which suits broadcasters and content owners with existing libraries.
How can a broadcaster build FAST channels from its existing library?
Point Media Tech's FAST (Kaster) reads your existing NAS via an Edge Agent; masters stay put and only 1–5% of derived data goes to the cloud. AI semantic search helps select content, and rule-based break scheduling generates the EPG automatically. Channels play out on a local playout engine that keeps running if the cloud link drops.
How do FAST channels get onto FAST platforms?
Point Media Tech's FAST outputs HLS streams and EPG data (JSON / XMLTV) to each platform's delivery spec, so one schedule feeds several FAST platforms. Platform-specific specs, multiple time zones and loops are set in the scheduling module. Actual launch still follows each platform's own partnership and review process.
How do FAST channels insert ads with SCTE-35 and SSAI?
Point Media Tech's FAST emits SCTE-35 signaling from your break rules (window, interval, duration, position), then SSAI stitches ads in at the stream layer. Ads and program share one stream, so client-side ad blockers have nothing to block. For break signaling and server-side stitching in detail, see SCTE-35 Ad Insertion and SSAI server-side ad insertion.
FAST channels vs. paid OTT subscriptions: which model suits mid-sized media companies better?
Mid-sized media companies with deep existing libraries that want a low-barrier way to reach new audiences usually do better starting with FAST. Viewers watch for free and the channel earns through ad revenue share, with no subscription system to build. Paid OTT suits platforms with an established subscriber base and budget for membership and marketing; see white-label OTT platform.
What does a FAST channel cost, and how is licensing calculated?
Point Media Tech's FAST is licensed by channel count and feature modules, with cost depending on output channels and ad-insertion needs. Plans are monthly or annual; it deploys in containers in the cloud or on premise, and the Edge Agent runs on your existing NAS or an x86 host with no GPU purchase. Contact sales for a quote, or start with a 100-hour POC.
Ready to monetize your library
24/7 — without moving it?
Book a demo, or request a 100-hour POC, and see how Point Media Tech's FAST schedules your existing library into FAST channels playing out 24/7.