Why Artist Alley

The problem

Studios that ship digital art — games, animation, illustration, motion design, VFX — review and archive their work across Slack, Miro, Teams, Drive, Dropbox, Frame.io, WeTransfer, and whatever shared folder a producer set up three years ago. Knowledge fragments across tools. Review is inconsistent. Versioning is improvised. Assets become hard to find a few years later, and the institutional memory of a project evaporates when contributors move on.

The market has two kinds of tools available:

• Open-source asset managers that optimize for archivists. They were built for libraries and museum collections; the artist’s workflow is an afterthought, and the review experience hasn’t been touched in a decade.
• SaaS review platforms that nail the review loop but lock everything behind a subscription, only ship the slice they sell, and own the storage. None are self-hostable. None speak the production language of game and animation pipelines.

Neither solves the full problem. Studios end up stitching them together — and paying for several at once.

And there’s a third gap nobody is filling: neither category federates. When a studio needs to share work with an external partner — a co-development studio, a marketing agency, a licensing partner, a museum on loan — the answer today is either a Frame.io seat per outsider or a Dropbox link with the metadata stripped off. Federated social networks exist — Mastodon, PeerTube, Pixelfed. Federated DAMs don’t, as far as we’ve found.

Artist Alley is built to be the single self-hostable system that covers the full loop: upload, organize, review, version, archive, share, ship. Production tool and archive, by the same artists who use it day to day, with federation between independently-operated studios as a first-class capability rather than an afterthought.

The federation protocol — ArchivePub — is open, ActivityPub-shaped, and implementable by anyone. We ship the reference implementation. Federation between studios shouldn’t be something you rent from a vendor; it should be a property of the protocol.

Who it’s for

The product was shaped around the game and animation studio pipeline, and the feature set reflects that: a review canvas, frame-by-frame playback, native 3D viewers, a brand workspace, and federation across teams — with NDA gating, presentation rooms, and DCC integrations on the roadmap.

Game and animation aren’t really separate verticals. A studio’s cinematics team reuses the in-game art; the game team pulls finals from cinematics; brand owns the bible both pull from. The asset flow is bidirectional and the tooling is the same. Sub-specialisations follow — concept, modelling, rigging, animation, FX, lighting, comp, brand, cinematic editorial — but the underlying object (asset → review → version → archive → ship) is one model.

The model generalises beyond production studios. Anywhere a team catalogs items with associated digital media, Artist Alley works:

• Museums tracking physical artifacts with provenance scans, measurement data, condition reports.
• Libraries digitising manuscripts, books, periodicals with their derivative metadata.
• Galleries managing inbound and outbound loans with custody chains and verification photos.
• Archives keeping document collections with scan provenance.
• Studios managing physical props, scripts, costumes, set pieces — each with their digital companion files.

The feature set is tuned for production studios first; the foundation is general — and physical-archive operators are slated to get explicit first-party support through the physical_archive feature flag (Phase 1.51, ADR 0048). Flipping the flag will turn on per-collection accession-number formats with auto-increment, a hierarchical location tree (building → room → cabinet → shelf → bin), a loan-management system (loans-in and loans-out with the standard 7-state workflow + insurance + signed-agreement attachments + condition reports before and after handoff), provenance chain-of-custody tracking, and standards interop (Spectrum 5.0 procedure crosswalk, LIDO XML import/export, Dublin Core projection, CIDOC CRM crosswalk document, optional Getty Vocabulary auto-complete). The flag is default-off — game and animation studios never see the additional surfaces. Museums, galleries, libraries, archives, and studios with prop archives will flip it on at install time.

By role:

• Artists — upload work, see review notes, version iterations, find the brief.
• Reviewers + leads — work through unreviewed assets and leave async feedback on a zoom / pan / tile canvas (live presenter sessions land with the review-tool arc).
• Producers + studio operators — stand up a single binary, point it at Postgres, done. Audit, permissions, share links, retention, federation across teams all in one surface.

How it’s licensed and priced

Artist Alley is open source — AGPL-3.0-only, dual-licensed with a separate commercial license (ADR 0016). The core ships with no paywall — the features we ship are in the open-source binary an operator can self-host forever.

The commercial path matures alongside the product. The intended shape: capacity tiers for studios at scale (seat / asset / project ceilings rise with the tier; the AGPL core stays feature-complete at every tier) and premium add-ons for capabilities that either let operators monetize their own users (commerce, ads, memberships, paid storefronts) or carry real operational burden on our behalf (cloud-hosted AI, premium SSO bridges, managed backup, paid DCC plugins). Locally-run alternatives stay free in every case.

Specifics — pricing, seat counts, tier features — get nailed down at Phase 1.24 when the licensing infrastructure ships. For now: install it, run it, use it. See ADR 0017 and ADR 0038 for the architectural posture.

What we explicitly aren’t

• A general-purpose digital asset management product for archivists. We’re shaped for studios that actively produce art. The archive is a byproduct of the production workflow, not the starting point.
• A pipeline integration platform. We’re a destination for finished and in-progress art, not a Perforce / ShotGrid / Houdini replacement. Importing from those tools is on the roadmap (Phase 1.43); we don’t replace them.
• A hosted-only SaaS product. Self-hostable by design. A managed hosted offering may follow for indie studios without IT capacity, but Enterprise customers always self-host.
• NFT / crypto / blockchain anything. Different audience, different value system. Not for us.

Non-negotiables

• Single-binary deploy. One Go binary. SvelteKit SPA embedded via go:embed. Configure with environment variables; that’s the install.
• Three production containers. nginx, app, postgres. Heavier add-ons (Blender worker, AI models) ship as optional companion containers.
• Postgres-only. No MySQL adjunct, no separate search index, no separate vector store. One place to back up.
• Open by default. The AGPL core stays feature-complete. Premium add-ons are additive, never clawing back something the tier already includes.

Read on:

• Architecture — how the binary, database, and frontend fit together.
• Design pillars — the principles we use to make product decisions.