Identity, sessions, entitlements, and usage aggregation
Own product access, session/device rules, license and key authority, audit, and billing integration.
Technology
One media engine. A composable platform.
CoreMF owns media mechanics. WMPlatform supplies replaceable providers. WaveMinds services own product identity, sessions, entitlements, and commercial authority.
Architecture
Clear boundaries from runtime to product.
Embed CoreMF by itself or connect it to the platform providers and independent services needed by your deployment.
Composable plugins and reusable services
Logging, security, policy, DB/event sinks, application database, media stores, index, AI, load balancing, CDN, DRM, and stream crypto.
Graphs, protocols, codecs, GPU production, and media encryption
Run standalone without platform dependencies or load platform providers through the host/plugin boundary.
Media & GPU
A shared foundation for live, recorded, and composed media.
Build media workflows on CoreMF graph processing and CoreGL production paths.
Protocol & runtime
Ingest, transform, and deliver
Connect WebRTC, RTMP, SRT, and HLS workflows. Combine recording, transcode/remux, SFU/meeting paths, and media graph processing.
GPU production
Compose the output your product needs
Filters, overlays, lower thirds, watermarks, transitions, layouts, broadcast/meeting compositions, and audio-reactive graphics.
Protection mechanics
Keep media and authority separate
CoreMF encrypts media, handles manifest signaling, and runs stream encrypt/decrypt nodes. Platform services own long-lived keys, authorization, license decisions, and vendor credentials.
AI integration
Bring the model. Keep the media workflow.
Use the AI plugin layer for detection, tracking, OCR, segmentation, face embedding, pose, and customer models.
Compose inference with media and platform services
Run selected models locally, connect results to recording indexes and event metadata, and apply tenant policy, usage, and privacy handling around the workflow.
CoreMF’s registered Python AI processor pack broadens the algorithm foundation. Each selected algorithm still needs product integration and retained model/runtime evidence.
Validate the model for your use case
AI plugins are integration baselines. Production model packages, dataset-backed accuracy, runtime performance, privacy behavior, and deployment profiles are part of evaluation and release selection.
Review AI release scope ↗Storage & data
Put media bytes and application state in the right place.
Storage, indexing, event emission, and application records have distinct roles.
| Boundary | Responsibility | Integration implication |
|---|---|---|
| Media stores | Recordings, clips, thumbnails, and other media bytes | Choose local, object, or tiered storage profiles and validate access and retention. |
| Media index | Recording locators and AI/event correlation | Connect search and retrieval to the media your application needs. |
| Event/audit sinks | Fanout of append-only operational facts | Emit audit, detection, and usage-like events to selected sinks. |
| Application database | Queryable metadata and mutable workflow state | Postgres baseline exists; catalog/alarm APIs, migrations, and production hardening remain integration/release work. |
Integration
A stable plugin boundary and independent services.
Extend the platform without merging every integration into the media engine.
C++ and plugin SDK
WMPlatform shared objects load through a stable C ABI for plugin discovery, factory creation, and teardown. Providers connect your storage, identity, models, policies, and media protection to the runtime.
Independent service processes
The root CMake build covers CoreMF and WMPlatform plugins. Go services run independently with their own deployment and operational profiles.
Evaluate with a reproducible baseline
The repository includes a day-one demo, evaluation Docker image, plugin SDK guide, and internal verification paths. Agree on the selected configuration, dependencies, expected evidence, and operational runbooks before production rollout.
Discuss integration requirements ↗Start a conversation
Scope a platform profile around your workflow.
Tell us about your workflow, deployment constraints, and integration needs. We will scope an evaluation around the right package and deployment profile.