Active Titles
14
Across 4 disciplines
↑ 3 from last quarter
Avg Cycle Time
9.2mo
Target: 8 months
Was 17.4 months
AI Acceptance Rate
84%
Minor or no edits needed
↑ 6% this month
Derivatives Generated
2,847
Test items, slides, IM sections
This quarter
Handoffs Eliminated
63%
vs. pre-Forge baseline
Target: 60%
Active Title Pipeline
14 titles in progress
| Title | Phase | Progress | Risk | Cycle |
|---|---|---|---|---|
|
Introduction to Economics, 5e
Dr. Sarah Chen · Business & Economics
|
Derivative Gen |
72%
|
8.1mo | |
|
Human Anatomy & Physiology, 11e
Prof. Marcus Webb · Life Sciences
|
LD Review |
55%
|
11.3mo | |
|
Principles of Marketing, 8e
Dr. Anya Patel · Business
|
38%
|
On track | ||
|
Organic Chemistry, 4e
Prof. Jae-won Kim · Physical Sciences
|
Rights Clearance |
61%
|
At risk | |
|
Psychology: Core Concepts, 9e
Dr. Fatima Al-Rashid · Social Sciences
|
QA & Accessibility |
89%
|
7.8mo | |
|
Statistics for Business, 3e
Prof. David Okafor · Business & Economics
|
Planning |
12%
|
Scoping |
Exceptions Requiring Attention 3
Rights clearance blocked — Organic Chemistry 4e
7 figures from prior edition cannot be reused. Rights agent has identified 5 cleared alternatives. 2 require SME review.
Review alternatives →
Author SLA approaching — Anatomy 11e, Chs 14–16
Marcus Webb has not reviewed AI draft chapters. SLA expires in 3 days. Auto-escalation queued.
Send reminder →
Low AI confidence — Economics 5e, Ch 12 test bank
Quantitative items confidence at 71% (threshold: 80%). 14 items routed to human review queue.
Review queue →
Cycle Time Trend
8
Active agents
3
In configuration
1
Scheduled for deployment
84%
Portfolio acceptance rate
Narrative
🖊 Delta Detection Agent
● Active
Compares current manuscript against previous edition. Flags changed sections, identifies content requiring update, and generates a prioritised revision scope for authors. Eliminates blank-page problem.
94%
Accuracy
2.3hr
Avg per chapter
847
Chapters processed
Narrative
✍ Revision Draft Agent
● Active
Generates draft chapter revisions from delta analysis and author instructions. Maintains voice, style consistency, and editorial standards. Authors review and approve rather than create from scratch.
87%
Acceptance rate
4.1hr
Saved per chapter
312
Drafts produced
Quality
📝 Copyedit Agent
● Active
Runs copyediting passes against house style guide, grammar rules, and consistency checks. Operates continuously during authoring — not as a late-stage gate. Flags issues inline for author review.
96%
Style compliance
0.4hr
Per chapter pass
4,210
Issues flagged
Derivatives
📋 Test Bank Agent
● Active
Generates multiple-choice, short-answer, and essay questions aligned to learning objectives. Produces Bloom's taxonomy tags, difficulty ratings, and answer rationales. STEM content routes to SME verification.
82%
Acceptance rate
1,847
Items generated
71%
STEM confidence
Derivatives
🎯 Instructor Manual Agent
● Active
Updates instructor manual sections in parallel with narrative revision. Maintains alignment between source content, learning objectives, and instructor guidance. Eliminates separate manual update workflow.
91%
Acceptance rate
6 wks
Time saved per title
430
Sections updated
Derivatives
📊 Slide Deck Agent
● Active
Generates and updates PowerPoint lecture slide decks from structured source content. Maintains visual consistency, key concept coverage, and learning objective alignment across all chapters.
88%
Acceptance rate
570
Slides generated
3.2hr
Saved per chapter
Rights & Compliance
⚖️ Rights Classification Agent
● Active
Classifies rights status for all assets referenced in a revision. Identifies clearance requirements, cost flags, OER alternatives, and expired permissions. Surfaces issues at authoring stage, not post-production.
92%
Classification accuracy
1,204
Assets classified
4 wks
Avg time saved
Accessibility
♿ Accessibility Agent
● Active
Runs WCAG 2.2 compliance checks, generates alt-text for images and figures, checks colour contrast, and validates heading structure. Embedded in authoring workflow — not end-of-cycle audit.
98%
First-pass compliance
3,200
Alt-texts generated
Zero
End-cycle QA failures
Planning
📅 Scheduling Agent
⚙ Configuring
Automated cascading schedule generation from title parameters, resource availability, and historical cycle-time data. Replaces manual schedule building and maintenance. Deployment Q3 2026.
—
In config
—
In config
Q3
Deploy target
Workflow Transformation: Edition-Based vs. Forge Continuous Operations
❌ Before Forge — Edition-Based Production
Author revises in MS Word, uploads to Box. No visibility of what changed vs. previous edition. LD receives file by email.
12–18 mo
Sequential LD review → copyedit → rights → production. Each phase waits for the previous one. Multiple email chains and file transfers.
8 wks
Derivatives built separately. Test bank, IM, slide deck each updated manually against source content. Inconsistency common.
4–6 wks
End-of-cycle QA, accessibility, rights check. Errors found late. Expensive rework. Accessibility failures require layout redesign.
3–4 wks
Release. Edition locked. Next revision cycle begins from scratch.
Fixed release
→
✅ After Forge — Continuous Orchestration
AI delta detection identifies what changed. Author opens prioritised change list in the repository — not a blank page. AI drafts revisions for review.
Days
Parallel execution. LD reviews inline. Copyedit, accessibility, and rights checks run continuously during authoring. No sequential waiting.
1–2 wks
Derivatives auto-generated from repository. Test bank, IM, slides, courseware drafted by agents in parallel. Human review queues generated automatically.
1–2 wks
Shift-left quality — no end-of-cycle surprises. Accessibility, editorial compliance, and rights status resolved during authoring. Zero final QA failures.
Continuous
Continuous release capability. Content updated when it needs to be, not when an edition cycle permits. All derivative outputs automatically in sync.
On demand
Derivative Generation Efficiency — Introduction to Economics, 5e
⭐ Golden Content Repository — Source Title
Introduction to Economics, 5e
Chapter 12 — Supply & Demand
28,400 words
Chapter 13 — Market Equilibrium
24,100 words
Chapter 14 — Elasticity
22,800 words
Chapter 15 — Consumer Theory
26,500 words
Chapter 16 — Production Costs
23,900 words
Auto-Generated Derivative Outputs — Ch. 12 (Supply & Demand)
Test Bank
45 questions generated · 84% acceptance rate · 7 queued for SME review
38 MCQ
5 Short-answer
2 Essay
Conf: 88%
Instructor Manual
12 sections updated · 91% acceptance rate · Ready for LD sign-off
12 sections
4,200 words
Conf: 93%
Lecture Slides
34 slides generated · 88% acceptance rate · Visual review pending
34 slides
8 diagrams
Conf: 90%
Accessibility Check
WCAG 2.2 compliant · 28 alt-texts generated · 0 issues outstanding
28 alt-texts
100% compliant
Shift-left ✓
Hard Governance Gates
🔒 Copyedit Lock — Economics 5e
No content changes permitted once copyedit stage begins. Gate enforced: 19 May 2026.
✅ Version Lock — Psychology 9e
Committed content is immutable. Patch revisions via formal RFE only.
⏳ Scope Lock pending — Anatomy 11e
Awaiting author confirmation on Chapters 14–16 scope. Lock triggers on sign-off.
✅ Budget Lock — Marketing 8e
Budget variance >10% triggers escalation gate. Currently within tolerance.
🔒 Proofing Gate — Psychology 9e
Maximum 1 round proofing. No exceptions. Author SLA enforced by system.
AI Governance Requirements
Explainability logs for every AI decision. Every agent action attributed to a specific agent version, with decision context stored.
Versioned prompts and models. No silent updates to AI behaviour. All prompt changes logged and approved before deployment.
Human override at every AI decision point. No AI output progresses without explicit human sign-off. Override authority preserved.
Configurable confidence thresholds by workflow stage. Outputs below threshold route to human review — not auto-approval.
Defect rate by discipline is a hard gate. STEM/quantitative agents suspended if defect rate breaches agreed threshold.
Override log required. All human overrides of AI decisions captured, reviewed, and fed back into agent improvement cycles.
Audit Trail — Recent Activity Millisecond-precision logging · All actions attributed
09:41:22
AI Agent
Test Bank Agent generated 45 questions for Economics 5e Ch.12. 38 accepted (MCQ). 7 routed to SME review queue (confidence 71% — below STEM threshold of 80%).
Conf 84%
09:38:04
S. Chen
Author approved AI draft revision for Economics 5e Ch.12 with minor edits (3 sentence changes). Chapter marked LD-ready and submitted for inline review.
Human approval
09:31:17
AI Agent
Copyedit Agent completed pass on Economics 5e Ch.12. 14 style flags raised, all resolved. 2 ambiguous constructions flagged for author confirmation.
Conf 96%
09:14:53
AI Agent
Accessibility Agent: 28 alt-texts generated for Economics 5e Ch.12. WCAG 2.2 compliance confirmed. 0 outstanding issues. Shift-left check complete.
Conf 98%
08:57:30
M. Osei (LD)
Learning Designer approved Instructor Manual sections for Economics 5e Ch.12. Override logged: LD revised 1 section for pedagogical accuracy. Agent improvement flagged.
LD override logged
08:44:12
AI Agent
Rights Classification Agent: 4 figures from prior edition flagged for rights review. 3 cleared OER alternatives identified. 1 requires manual clearance — exception queue populated.
Conf 92%
08:30:01
AI Agent
Delta Detection Agent completed analysis of Economics 5e Ch.12 vs 4e. 34% content change detected. Prioritised revision scope generated. Author notified via workspace.
Conf 94%