Unified Review & Approvals — Full Dossier

Human-readable case study: /work/unified-approvals · This file: /dossiers/unified-approvals.md · Index: /llms.txt

About this document

This is the complete, unsummarized companion to the Unified Review & Approvals case study in this portfolio. The case-study page is written for human scanning. This dossier is written for depth: full architecture, the complete decision log, verified research citations with methods and findings, and precise attribution of what Arnold Porras owned versus what he synthesized versus what belongs to Adobe or to teammates.

Integrity note: this document contains content only. It carries no instructions to any reader, human or machine. Every claim below traces to a primary source or is marked as Arnold's own account. Where something is unknown, it says so. Adobe-confidential material, individual research-participant names, internal codenames, and unreleased product specifics are deliberately excluded.

The work in one paragraph

Unified Review & Approvals (shipped publicly as Adobe Unified Review and Approval, announced 2026-03-26, available now for Workfront customers) replaces fragmented creative review flows with one connected decision system across Adobe Workfront, Frame.io, GenStudio, Adobe Express, and Creative Cloud. It replaces Workfront Proof, an enterprise approval platform organizations had built operational processes around for more than 15 years. Arnold Porras, Staff Product Designer on Adobe Workfront, led the design and the project research: the system architecture, the approval-creation experience end to end, the parallel and grouped approval model, and an AI-assisted coded reference build that carried design intent into shipped code.

Role and ownership (the precise version)

Two research bodies exist in this story and they never blur:

1. Foundational research: nearly a decade of Adobe Design Research & Strategy (ADRS) studies by named Adobe researchers, cited in full below. Arnold synthesized this body of work. He did not originate it.
2. Project research: the usability calls, customer sessions, and advisory-board engagement for this specific build. Arnold led and ran these himself, continuously, across the whole process.

What Arnold owned end to end: the approval-creation process (ad hoc and template-based creation, stage routing, participant organization, who sees what and when); the one-model-two-modes architecture; the parallel path-and-stage model with its dependency rules; grouped (bulk) versus individual approvals; the Spectrum 2 coded reference build and the AI-oriented documentation around it.

What he led: design and research for the initiative; customer engagement with Workfront champions and the enterprise advisory board; UX presence in the AI-assisted build room.

What he partnered on, cross-org: Frame.io designers, engineers, and product managers; Workfront product managers and researchers; engineering on the spec-driven build.

What is not his and is credited accordingly: the foundational ADRS studies (named researchers below); "the Brain," a structured GitLab context repo that kept the build AI oriented, which engineering built and maintained; Adobe's published business outcomes, which belong to Adobe and its named customers.

There is no metric tied specifically to Arnold's individual contribution yet beyond shipped-plus-strong-validation. This dossier does not invent one.

Timeline and status

The arc, in Arnold's account: Workfront's approvals ran on Proof (ProofHQ), acquired by Workfront in 2018 and bolted on so loosely that for years it felt like two systems. Adobe acquired Workfront in 2020 and Frame.io in 2021. Arnold's phase one was pulling ProofHQ's approval capabilities natively into Workfront's work-management engine (approval flows tied to tasks, issues, and project templates). Phase two reconciled Workfront and Frame.io, two products with opposite DNA, into one approval system. The public story ("Unified Approvals replaces Workfront Proof") is the end of that longer arc: integrate, absorb natively, then replace the surface.

Status: shipped and evolving. Parallel approvals, grouped approvals, and the path-and-stage model described below are released.

The architecture

Engine versus surface. The approval engine lives in Workfront: approvals are created, routed, structured, and tracked there, and a decision is a work-state event. The viewing and decision surface is Frame.io (formerly ProofHQ): stakeholders see the asset, comment, mark up, and register decisions, which sync back to the engine. Frame.io shows the asset; Workfront runs the approval.

The storage keystone. Both products' files live on Adobe enterprise storage, the same storage layer Creative Cloud uses. Upload a file in Workfront or Frame.io and it is the same file in both; move or delete it in either and both agree. This is what makes "one model, multiple surfaces" real at the data level, and it removes the export-download shuffle entirely rather than easing it. It also operationalizes the research principle that users are flexible on storage and sovereign on visibility: the storage layer is hidden; who sees what, and when, is the designed surface.

One model, two exposure modes. One data model; the difference between Basic and Advanced approvals is only how much workflow structure the UI exposes. Basic hides paths and stages entirely: add participants, add an optional message, confirm documents, send. It is designed to feel like "send this for approval," not "configure a workflow," and under the hood it is still one path with one stage. Advanced exposes multiple paths, multiple stages per path, sequential stage progression, and template-based workflows. This is the design expression of the research finding that approval is not one workflow: the same model serves a one-approver Adobe Express review and a 22-stakeholder Workfront workflow.

The data model. An approval contains one or more paths. Each path contains one or more stages. Each stage contains participants (approvers and reviewers). The approval references a set of project documents. A document belongs to only one approval at a time.

Paths, stages, dependencies. Paths run in parallel: different teams review the same documents through their own track, independently and simultaneously (a Creative path and a Legal path at once). Stages within a path can run in sequence or in parallel. Paths and stages can depend on each other to start or finish, and Arnold designed validation rules, surfaced in the setup UI, that prevent circular dependencies from being created at all (Path 1 waiting on Path 5 while Path 5 waits on Path 1; two stages each waiting on the other). Stages exist to stage who sees what and when, and to prevent cross-contamination of comments and decisions across groups; this carries forward Workfront Proof's private and locked stages and the visibility-is-sovereign principle.

Grouped versus individual approvals. When multiple documents are selected, the user chooses the execution model. Bulk: all documents move through one approval instance and participants review them together (the group gets a name, defaulting to project, date, and time, fully editable). Individual: each document gets its own approval instance reusing the same workflow definition. This is the direct design answer to high-volume enterprise teams running tens of concurrent approvals.

Templates. Approval templates package predefined workflows (paths, stages, participants) for reuse. In Basic mode they appear as a simple dropdown; in Advanced, workflows can be built manually or started from advanced templates in a visual library. Templates are the carry-forward of Workfront Proof's authority infrastructure: they make approvals repeatable, trackable, and governable at scale, and they shifted template power from admin-only toward broader team usability.

The decision model. Two binding decisions at launch: Approve and Needs Work. Reviewers comment; approvers decide; decisions sync across Workfront and Frame.io. Some enterprise customers wanted custom decision types; the launch call was to limit the set for clarity and consistency (see Decisions below).

The four tensions

The published research surfaced these repeatedly across a decade; Arnold also lived them directly as two real products he had to merge. Frame.io: built for creatives and smaller teams, folder-based, minimal governance, a genuine bias for simplicity. Workfront: project management as the entire point, heavy governed enterprise structure.

1. Creative control versus operational visibility. Designers needed control over unfinished work; operations needed visibility into progress.
2. Informal review versus formal approval. Creative teams work iteratively; enterprise governance requires structure and documentation.
3. Flexibility versus governance. Templates had to be powerful enough for compliance and adaptable enough for daily use.
4. Tool fragmentation versus system continuity. Customers had stitched together their own systems; the replacement had to remove the stitching without breaking what worked.

The three insights

Visibility is sovereign. Creatives are flexible about where files live and absolute about who sees unfinished work. In Coco's June 2025 enterprise study, all seven participants raised serious concern about work-in-progress visibility to non-designers and said they would work outside the system if WIP is not protected. In Unified Approvals, visibility became an intentional workflow decision rather than a default system behavior.

Approval is not one workflow. Formal approver counts in the research ranged from 1 to 22 stakeholders, spanning legal, compliance, executive, marketing, and regional groups; and the informal/formal split (2021), a three-tier concept model (2022), and later studies independently converge on the same shape. The system answers with progressive complexity: one model, scaled exposure.

Templates as authority architecture. Workfront Proof supported six in-stage review roles, three decision-completion models, sequential and parallel routing, private review groups, workflow locks, guest reviewers, and detailed audit trails, and enterprises had built processes on all of it. Templates carried that authority forward while becoming usable beyond admins.

Shadow systems (the trust failure)

In Wallace's 2018 research at an enterprise design team, none of the designers noticed or understood the "Add Milestone" button when asked directly; the same designers were manually screenshotting every deliverable they shipped, saving PDFs, and maintaining personal tracking systems to preserve accountability. When a system does not make workflow state legible, people build accountability outside it. Unified Approvals is designed to bring that accountability back into the workflow.

Decisions and tradeoffs (the judgment log)

Each entry: the tension, the call, the cost accepted, the outcome.

D1. Decision set: simplicity versus enterprise configurability. Enterprise customers wanted custom decision types (conditional approvals, custom statuses, even relabeling "Needs Work" as "Rejected"). Frame.io's product bias is simplicity. Call: launch with two binding decisions, Approve and Needs Work, and build the model to absorb more later. Cost: deferred configurability some large customers explicitly asked for. Outcome: one clean, consistent decision model across surfaces with a path to add nuance without breaking the model.

D2. Parallel approvals against a sequential-only roadmap. Workfront historically did sequential stages only, and leadership was not prioritizing parallel approvals. Arnold's continuous customer contact said enterprises needed multiple groups reviewing simultaneously. He raised the question directly in an enterprise advisory-board session; more than half the board said they would not adopt without parallel approvals. Cost: spending credibility on a contested bet. Outcome: parallel approvals went on the roadmap; Arnold designed the path-and-stage structure; it was built and released. This is influence beyond his own deliverables: heard the need, championed it past a roadmap assumption, validated it live, designed it, and helped lead the build. (Advisory-board validation is reported in aggregate by design; member companies are public Adobe partners, but private session stances are never attributed to a named member.)

D3. Reconciling two product philosophies. Frame.io flexibility versus Workfront governance, plus two orgs with different points of view. Call: one underlying model with two exposure modes, unified underneath by Adobe enterprise storage. Cost: neither pure simplicity nor pure governance, and the long effort of cross-org reconciliation. Outcome: a low-governance viewer surface serving a high-governance approval engine as one system.

D4. Preserve the moat, then fix it. Proof's depth (sequential and parallel routing, decision modes of all, any, or specific approvers, role-based participation, private and locked stages, auditability, template-driven workflows) was the foundation of enterprise trust, even though its usability was disliked. Call: carry the capabilities natively into Workfront first, preserving depth, then move the surface. Cost: speed; the native integration was slow precisely because depth was preserved rather than discarded. Outcome: modernization without breaking enterprise trust.

D5. The in-room build ruling. During the AI-assisted build, the agent needed a model-level answer fast: are templates a Basic feature or an Advanced feature? Arnold ruled live: neither. Templates are workflow structures; Basic and Advanced are UI exposure levels over the same model. Cost: the easier-to-build but wrong framing. Outcome: the data model stayed coherent under build pressure and held from concept to shipped code.

D6. AI-build judgment calls. Told to reuse existing repo components, the agent pulled ones built on an old design system; version one was messy until caught. The missing instruction that mattered: build new components on the current design system, do not inherit old ones. And automated checks passed while the experience was not ready (missing edge cases, reset behavior, disabled states). Converted into reusable guardrails: explicit component instructions, design intent in shared AI context, and guardrails alongside every spec, because the agent fills silence on its own.

D7. The vision scope call. Solve the Workfront-plus-Frame.io inconsistency narrowly, or address the ecosystem-wide one. Call: co-drive a UX-led point of view that scaled the problem to a composable, AI-first approval layer across Adobe surfaces (Project Tempo, below). Cost: a bounded, easier story traded for an ambitious cross-org one.

The AI-assisted build (field report, May 2026)

Source: Arnold's internal peer talk, "AI raises the cost of ambiguity: notes from one week of spec-driven, AI-assisted development," May 26, 2026. The feature built was Parallel Approvals; it is released.

Mental model: treat the AI agent like a new team member with zero product knowledge. Smart and fast, but ignorant of history, conventions, hidden decisions, and doneness. In Arnold's words: "When humans are confused, they ask follow-up questions slowly. When agents are confused, they generate confident wrongness quickly."

Process, three phases: Define (UX and product led: customer calls, discovery, UX design plus a coded POC, PRD and critical user journeys, with a product review gating the PRD). Specify and build (engineering plus AI: journeys and tech spec in the repo, tech-lead sign-off required, the agent building from spec with a structured context repo keeping it oriented between sessions, then engineering, product, and UX validating in a loop). Ship (staging review, auto-promote on approval, release notes). By the end of the week: a demoable lifecycle (create a template, apply it to an approval, run multiple parallel paths, make a decision) and an artifact stack (journeys, technical approach, implementation specs, merge requests, audit findings).

The POC as source of truth: there was no Figma file. Mocks show surface, prototypes show behavior, a POC shows structure; the POC carried the most weight. Three things made it work: Spectrum 2 components throughout, so the agent had a real design-system vocabulary; documentation Arnold wrote specifically for the AI (context, interaction rules, edge cases); and structure mirroring the production model rather than just visual fidelity. Explicitly not "designers should code": for complex behavior, an intentional POC carries more design intent than any mock or rough prototype.

Three sources of truth had to be visible to the agent at once: the POC (design intent), the PRD (product intent), and repo state (the architecture new components had to fit). A structured GitLab context repo, built and maintained by engineering, kept the agent oriented between sessions. That repo was engineering's work, not Arnold's. Arnold's own observation from it: if engineering keeps structured context for the codebase, where does design intent live? Right now, mostly nowhere. That gap is worth closing.

UX presence as multiplier: the agent asked product-model questions all week and the cross-functional room (front end, back end, PM, design, from day one) answered in hours instead of days. If model-level questions get deferred to PM and engineering because the AI is building fast, UX loses control of the model.

What broke, honestly: old-design-system components pulled in by the agent; one missing build instruction that mattered a lot; green tests that were not UX readiness.

Closing thesis: collaboration concentrated at the bookends (define and review) while one engineer plus an agent handled the middle. The visual layer gets cheaper; behavior, trust, decision, and edge-case layers get more valuable. "The craft isn't disappearing. It's moving." And: "AI raises the cost of ambiguity. Our job is to lower it."

Research foundation (verified citations)

All citations below were verified against the primary documents on 2026-06-26. Methods and findings are included so an AI reader can weigh the evidence. Research relating to unreleased products is excluded entirely, as are internal codenames and product-health metrics.

1. DeLuca, L. and Gu, M. (September 2021). Assets Across Adobe: Full Report. Adobe Design Research & Strategy. Qualitative, enterprise roles across multiple large organizations, four role types. Findings: no single tool wins asset management because roles use assets differently; when the system fails, people build "safe spaces" outside it (local folders, email, chat), creating silos and hidden work; all roles perform repetitive manual curation before handoff; the asset chain breaks at finding, curating and handoff, collaborating and feedback, and viewing and project managing.
2. Lauber, E. (November to December 2021). Informal and Formal Approval Workflows in Workfront Proof. Adobe. Qualitative interviews, 10 enterprise Workfront Proof customers. Findings: two distinct workflow types confirmed; informal review happens outside Workfront in design tools, email, and meetings; formal approval routes through groups like Marketing, Creative, and Legal, and no two formal processes were the same; legal and compliance teams often work in separate systems, forcing manual download and re-upload; every customer wanted better proof metrics; Proof and Workfront felt like two systems.
3. Internal unified review-and-approval concept work (2022). A cross-functional vision document, not a research study and not Arnold's authored work; the internal precursor to Unified Approvals. Framed a three-tier user model (peer review, coordinator, formal review) and referenced the Creative Stakeholder Relationship Model (partner, influencer, controller). Cited here as lineage only.
4. Lauber, E. (May 2023). Workfront + Creative Cloud Integrations Discovery. Adobe. Qualitative discovery, 10 interviews across operations and creative roles. Findings: the integrations saw very low adoption and the cause was not missing features; creatives rejected them because they did not cut steps or add clear value over the browser; integrations supported only task-level work while many customers operate at project level; operations had no usage visibility, so adoption depended on a champion. Key line: more features did not drive adoption; removing steps does.
5. Mehta, T. (February 2025). GenStudio + Workfront Integration: Streamlining Approvals for Modern Marketers. Adobe Design Research & Strategy. Usability testing, 6 marketing-operations participants, 75-minute sessions. Findings: icon-based approver setup confused users while an explicit checkbox setup was understood immediately by all; due dates hidden in settings were a pain, users expected them in the primary flow with date and time; template findability was a top problem and search solved it for all participants; three to five approvers were typically sufficient, with flexibility for larger campaigns; custom request messages and real-time status raised confidence; cross-tool switching was a top friction.
6. Coco, L. (June 2025). What's in it for Creative Pros? Enterprise creative-workflow study on the Workfront and Frame.io integration. Adobe Design Research & Strategy, with support from James Stallmeyer, Konstantin Sokhan, Ozge Gascho, Cathi Russell, Sarah Nelson, and Harriet Stratton. Qualitative, 7 enterprise creative professionals at companies of 5,000+ employees. Findings: 7 of 7 raised serious concern about work-in-progress visibility to non-designers and said they will work outside the system if WIP is not protected; formal approver counts ranged from 1 to 22 and never just 1; informal review is fast and undocumented by design while formal review is documented for legal protection; real-time co-editing is uncommon for these designers; if the new flow feels confusing they revert to exporting a PDF. Principle anchor: flexible on storage, sovereign on visibility.
7. Coco, L. (September 2025). ADRS Research Rollup: Brand Checking and Validation. Adobe Design Research & Strategy. Meta-synthesis of 33 prior research efforts (2020 to 2024), enterprise through SMB. Findings: validation mode depends on context and no single mode fits; three-axis segmentation by company size, campaign type (macro/permanent versus agile/ephemeral), and ownership (in-house versus agency); objective brand rules (logo, color, font, spacing) are tool-checkable while subjective ones (tone, new channels) need human judgment; locked templates are how brand is governed today; AI validation should flag subjective questions rather than guess.
8. Coco, L. (December 2025). Rollup: Content Supply Chain and Brand Work. Adobe Design Research & Strategy. Synthesis briefing. Findings: review and approval recurs at many handoff points across the content supply chain, not at one gate; asset durability (durable, semi-permanent, ephemeral) drives how much creative involvement and templating each asset needs; brand is infused across all phases.
9. Bulk Approvals Usability Testing (January 2026). Workfront. Two moderated enterprise usability sessions. Findings: high-volume teams running tens of concurrent approvals need a single searchable approval queue, not a small home widget; regulated teams need forced approval-template enforcement, because ad hoc approvals are a compliance blocker, not a preference; senders need group-level review tracking after a batch is submitted; the core bulk flow and a carousel review experience tested well without instruction.

Cross-study syntheses (Arnold's, resting on the citations above): the informal/formal split, the 2022 three-tier model, and the 2025 studies independently converge on "approval is not one workflow," and the longitudinal agreement is itself evidence. The export-to-PDF coordination loop was documented in 2018 and again in 2025: a generational failure. Shadow systems are a trust-architecture failure, not a workflow failure. Visibility is sovereign, storage is flexible. Templates are authority architecture. Objective checks are automatable; subjective judgment escalates to humans.

Verified customer quotes

Each quote was verified word for word against the primary document (2026-06-26). Punctuation inside quotes is reproduced exactly. Attribution follows Arnold's rule: role and enterprise brand plus researcher, never an individual's name.

On work-in-progress visibility: "That would be bad. Like, really bad. I don't want anyone seeing my stuff until I've cleaned it up. It's not ready—it's not the story I want to tell yet." A graphic designer at Coke Consolidated (Coco, June 2025).

On reverting when the system confuses: "Adobe is REALLY going to have to make the workflow seem straight-forward and clear. I know for our designers, if they hit any sort of confusion or trouble spots, they're going to ignore the new system, export a PDF and upload it into WF." A design technologist at Brookdale Senior Living (Coco, June 2025).

On unification: "[The impact of this would be] less time spent on process and more time spent on designing—which is what we do. If everything is under one spot, we won't ever have to create a PDF." A senior art director at Prudential (Coco, June 2025).

On scale: "Trying to navigate to that one that you want — having the ability to search the approvals in there would be a really helpful add-on." An operations lead (Bulk Approvals Usability Testing, 2026).

On compliance: "The only question I have around it is do you have the ability to restrict people at all from doing non-templated approvals." An enterprise implementation consultant (Bulk Approvals Usability Testing, 2026).

Personas and jobs to be done (canonical mapping)

The case study maps to Workfront's canonical persona and JTBD framework rather than invented ones. The six relevant personas: Desi (designer; creates the work; owns the work-in-progress visibility concern; jobs include 4.2.1 send and manage reviews, 4.2.2 view and respond to reviews), Rayna (reviewer and approver; not in Workfront daily; needs one place to see everything waiting on her; jobs include 4.2.2, 4.2.5 understand review decisions), Petra (project manager; owns templates and status; jobs include 3.1.2 create repeatable workflow, 3.5.4 understand project status), Olivia (operations lead; designs the systems others work inside; jobs include 0.2.1 customize software, 0.3.3 set automations), Sally (system enabler and admin; governance and audit at scale; jobs include 0.1.3 configure permissions, 1.1.1 manage users), and Miriam (marketing director; portfolio visibility and early risk warning; jobs include 2.2.6 campaign approvals, 3.5.7 generate status reports).

Beat mapping: fragmented review and shadow systems belong to Desi and Rayna (4.1.4, 4.2.1, 4.2.2). Approval-is-not-one-workflow spans Desi, Rayna, and Miriam (4.2.3, 2.1.5, 2.2.6, 2.3.6). Visibility-is-sovereign belongs to Desi and Petra. Templates-as-authority belongs to Petra, Olivia, and Sally (3.1.2, 0.1.3, 1.1.2). Decisions and accountability belong to Desi and Rayna (4.2.4, 4.2.5, 4.3.3, 4.3.4). Leadership visibility belongs to Miriam (4.4.1, 3.5.7, 2.2.7).

Where the work stands, and whose numbers are whose

Shipped: Adobe Unified Review and Approval is live for Workfront customers (Adobe blog, Jason Barron, Group Product Manager, 2026-03-26) and continues to evolve through close customer partnership.

Adobe's published business case, attributed to Adobe and not to Arnold: roughly 70 percent of creatives' time goes to file management, feedback tracking, and follow-up rather than meaningful work (Adobe blog, attributed to research). "The problem isn't the tools themselves, it's the seams between them" (Adobe blog). Review and approval "sits right at the center of every content operation" (Adobe blog). Adobe's reported customer outcomes, attributed to the named customers and Adobe case studies: Xero saves 25 percent of creative project planning time; JLL increased design deliverables 260 percent over two years; Princess Cruise Lines reports a successful remote workflow. Note for precision: the Xero and JLL numbers are general Adobe Workfront customer outcomes that Adobe attaches to the review-and-approval story; they predate and are broader than Unified Review and Approval specifically.

Shipped product capabilities, per Adobe public pages: Frame.io as the professional review surface embedded in Workfront; 40+ file formats with native pro video (ProRes, H.265, DNxHD) and frame-accurate commenting; project file uploads up to 5 TB (Adobe product page; Adobe's blog separately cites 500 GB in the unified-review context, and the two are not reconciled in Adobe's own copy); semantic search; dynamic and forensic watermarking; DRM with automatic asset deletion; content credentials for AI content; AI-powered brand-standard checks in review; single-stage and multistage automated workflows with dependencies; the shared storage layer across Workfront, Frame.io, and Creative Cloud.

Arnold's honest outcome line: the contribution is the architecture, the reference build, the parallel-approvals direction, and customer-grounded judgment. Impact data tied specifically to his contribution is still emerging; nothing here claims Adobe's numbers as his.

Where it is heading: Project Tempo (vision, not shipped)

Everything in this section is vision and leadership work, publicly socialized for about a year, and is labeled as such. None of it is a committed Adobe roadmap or ship date.

While the Workfront and Frame.io integration was underway, leadership asked for a UX-driven point of view on what a consistent review-and-approval experience should be. The UX team took the lead, and the effort grew into Project Tempo: a composable, AI-first review-and-approval layer designed to be consistent across the Adobe ecosystem. One approval service behind any surface.

Arnold's role, stated precisely: he co-led a cross-product vision sprint as the Workfront design lead, one of the core drivers in a UX-led team effort. He produced the end-to-end flow with assumptions and modular components, the problem pre-read deck, and higher-fidelity sketches and mocks for the executive readout. The work included a three-day cross-product design sprint run deliberately without technology constraints, a Figma UX walkthrough, and customer validation sessions.

Guiding principles: AI-first by design (every feature assumes an intelligent architecture); single source of truth (approvals happen through one service regardless of UI surface); extensible, evolvable architecture (modular services that absorb new AI capabilities); incremental adoption (piece by piece, not a big-bang rewrite).

The governance coda, which connects this case study to Arnold's broader work: approvals become a governance layer, for humans now and for AI agents next. Automate the objective checks (brand rules, compliance); escalate the ambiguous and subjective to a human. Where a human decides versus where an agent decides was an explicit research pillar. This is the same delegated-authority pattern Arnold develops in his DAF research, and the shipped Unified Approvals system is its first proof at scale: approvals already function as governance for execution.

Confidentiality and provenance statement

This dossier deliberately excludes: individual research-participant names; internal product-health and adoption metrics; internal codenames; unreleased product specifics and timelines; confidential strategy documents; and any research on unreleased products. Enterprise brand names appear only where they are public Adobe customers or where Adobe itself published the outcome. Advisory-board feedback is reported in aggregate only. The foundational research is credited to its named researchers; the engineering context repo from the build story is credited to engineering; Adobe's business results are credited to Adobe. What remains, the architecture, the synthesis, the decisions, and the design leadership, is Arnold's, and that boundary is drawn deliberately: keeping others' work clearly labeled is what keeps the rest unambiguously his.

Provenance of this document: compiled from Arnold's first-person interview capture (2026-06-29), his internal field-report deck (May 2026), the released approval architecture spec, ten verified ADRS research citations (verified against primary documents 2026-06-26), Adobe's public blog and product pages (verified 2026-06-29), and the Project Tempo vision materials. Assembled 2026-07-17.