Manifold Grace

Projects & Architecture

A register of Manifold Grace software ideas, existing demonstrations and venture opportunities, with a baseline and an acceptance test for each main build.

AI model architecture families · 1,000 proposed configurations · Modular platform architecture · Assurance pathways · Academic crosswalk · Project register · 2027 opportunities · Learning outcomes · Related work

AI and machine-learning development — distinct types

One identifiable architecture family, algorithm or system method per project. The current four-level register distinguishes 17 active application and capability categories, 10 learning paradigms, 36 methods and frameworks, and 53 model or system-architecture records. Nine historical entries remain linked to their replacements. A model-architecture family, a system-integration pattern and a specific model topology are explicitly labelled as different kinds of Level 4 records. The current v1.6 register holds 116 active entries, adding nine historically documented search, symbolic planning, optimisation and graph-modelling approaches. SRC-009 provenance and v1.4 scope descriptions have also been reconciled. Coverage remains incomplete and project implementations have not been independently verified.

Browse corrected four-level AI taxonomy · Earlier three-table classification · Architecture types · Methods and learning paradigms · Linked projects and paper initiatives · Matching Excel-compatible table (v1.6) · Original scenario research

Architecture, legal and security assurance learning pathways

Linked project specifications for studying original papers and permitted code sources, reconstructing architecture for teaching, testing compliance and cybersecurity safeguards, and choosing appropriate academic or professional paper formats. Fifteen candidate project families and 18 publication formats are distinguished, with beginner, intermediate and advanced pathways; proposals are not described as completed implementations.

Explore architecture, assurance and research project pathways · Paper and publication options · Difficulty and repository organisation

Interdisciplinary creative AI, social research and humanities

Eight further proposed project families (16–23) connect software and model architecture with aesthetics, artistic quality, human dignity, persuasion, social responsibility, ESG, creative labour, historical culture, safeguarding and regulatory requirements. Assess technical performance, subjective reception, social harms and governance evidence as distinct questions rather than as one unqualified score.

Interdisciplinary project pathways, methods and paper types · Eight collapsible research questions

Live archival reference implementation and three research papers

A reusable source-linked publication archive generates daily, weekly, monthly, annual and complete PDF catalogues, alongside accessible browser navigation and JSON/CSV records. The implementation indexes dated public site records rather than copying external articles. Three separately labelled working papers address design-science architecture, legal-policy options and the use of living evidence archives in higher education and advanced AI evaluation.

Explore daily-to-annual PDF archive · Research papers and source-code lineage · Reusable generator

Academic crosswalk

Read each row from left to right: project, syllabus and reading, programme outcome, then the specific application a reviewer can inspect. Project links lead to the existing source pages. The exercises describe evidence to produce or review.

Sources and interpretation

Weekly topics and reading assignments come from the History and Foundations knowledge map and selected reading audit. Cross-week applications borrow a reading from another point in the syllabus. Outcome numbers follow the official 2024/25 programme specification. The links are proposed alignments for academic review. LO8 needs independent research evidence and LO9 needs documented collaboration.

Project and practical taskSyllabus and readingProgramme outcomesPrecise intersection
Independent ChatGPT-style interfaceRun one multi-turn chat, export it, and test cancellation and unsupported devices.Week 1 · Thinking machinesTuring, Computing Machinery and Intelligence (1950)LO1 AI foundations and societal impact; LO7 Communication across audiencesRun identical prompts against two models and discuss whether fluent responses support a claim of intelligence.
Business process automationRoute an enquiry through approval and replay a duplicate event.Week 2 · Cybernetics and grounded cognitionWiener, Cybernetics (1948)LO1 AI foundations and societal impact; LO3 Case analysis and solutions; LO4 Applied ethical reasoningTrace how acknowledgement and error feedback change the next workflow action.
CRM and Salesforce-style pipelineMove a synthetic lead from enquiry to opportunity with role checks.Week 5 · Data as infrastructureMuldoon et al., A typology of artificial intelligence data work (2024)LO1 AI foundations and societal impact; LO2 Ethical and societal evaluation; LO3 Case analysis and solutionsIdentify who supplies, labels and corrects lead data, then test who can amend a mistaken classification.
ERP and resource planningTrace a test order through stock, invoice and reconciliation.Week 2 · Cybernetics and grounded cognitionWiener, Cybernetics (1948)LO1 AI foundations and societal impact; LO3 Case analysis and solutionsShow how a stock discrepancy feeds back into reconciliation and blocks an incorrect fulfilment decision.
Compliance and continuous AI assuranceLink a control to evidence, a failed test, remediation and sign-off.Week 9 · Design and accountabilityEubanks, Automating Inequality (2018), ch. 3 (cross-week application)LO2 Ethical and societal evaluation; LO4 Applied ethical reasoning; LO5 Policy and regulationIntroduce a failed control and trace who can see it, challenge closure and approve remediation.
Evidence verification and contradiction APIBenchmark claim coverage, contradiction and justified abstention.Week 3 · Expert systems to ChatGPTMinsky, Steps Toward Artificial Intelligence (1961) (cross-week application)LO1 AI foundations and societal impact; LO3 Case analysis and solutions; LO8 Independent researchSeparate explicit evidence rules from model judgement and compare errors on the same claim pairs.
Research discovery and literature-gap auditCompare retrieval against a fixed literature set and test one proposed gap.Week 5 · Data as infrastructureCrawford, Atlas of AI (2021), Data chapterLO6 Critical discussion; LO8 Independent researchRecord corpus selection and omissions, then check whether an apparent gap survives a broader search.
Cross-conversation context and private memoryFind a consented record, apply a correction and verify deletion.Week 5 · Data as infrastructureCrawford, Atlas of AI (2021), Data chapterLO1 AI foundations and societal impact; LO4 Applied ethical reasoning; LO5 Policy and regulationTest whether consent, correction and deletion affect which past records retrieval can surface.
Historical as-of retrieval and correction propagationReconstruct a claim at two dates and propagate a superseding record.Week 9 · Design and accountabilityWinner, Do Artifacts Have Politics? (1980) (cross-week application)LO3 Case analysis and solutions; LO5 Policy and regulation; LO8 Independent researchCompare visible record histories before and after a correction and trace who can authorise supersession.
Document-drop classifier and evidence organiserClassify mock files and preview filenames before applying any change.Week 5 · Data as infrastructureMuldoon et al., A typology of artificial intelligence data work (2024)LO1 AI foundations and societal impact; LO3 Case analysis and solutionsLog classification decisions, review disagreements and quantify the manual work behind reliable labels.
Contributor provenance and innovation attributionRecord origin, revisions and an attribution dispute.Week 5 · Data as infrastructureMuldoon et al., A typology of artificial intelligence data work (2024)LO4 Applied ethical reasoning; LO5 Policy and regulation; LO9 Interdisciplinary collaborationRecord who created, annotated and corrected a contribution, then test a disputed attribution.
Agency.ai agent workspaceRun a simulated service request with a reviewable action ledger.Week 9 · Design and accountabilityWinner, Do Artifacts Have Politics? (1980) (cross-week application)LO1 AI foundations and societal impact; LO4 Applied ethical reasoning; LO5 Policy and regulationDemonstrate how approval and override controls allocate authority between agent and reviewer.
Public-interest research and societal resilience adviserCompare two policy scenarios and cite assumptions and evidence gaps.Week 7 · AI, power and the stateEubanks, Automating Inequality (2018), ch. 3LO2 Ethical and societal evaluation; LO6 Critical discussion; LO7 Communication across audiences; LO8 Independent researchCompare who benefits from two policy scenarios and identify people the input data leaves out.
UK public bodies and oversight directoryTrace a body's remit and oversight across a dated reorganisation.Week 7 · AI, power and the stateWinner, Do Artifacts Have Politics? (1980)LO5 Policy and regulation; LO7 Communication across audiences; LO9 Interdisciplinary collaborationTrace an oversight handoff and show how the directory represents institutional authority over time.
Learning outcome, viva and competence verificationCompare rubric results on matched sample responses and collect reviewer agreement.Week 8 · Fairness and counter-imaginariesFerrara, Fairness and Bias in Artificial Intelligence (2023)LO2 Ethical and societal evaluation; LO3 Case analysis and solutions; LO8 Independent research; LO9 Interdisciplinary collaborationTest whether equivalent answers receive comparable rubric scores across writing styles.
Prior learning and skills evidence portfolioMap one sample portfolio to criteria with a human review decision.Week 8 · Fairness and counter-imaginariesKazansky & Milan, Bodies not templates (2021)LO4 Applied ethical reasoning; LO7 Communication across audiences; LO9 Interdisciplinary collaborationReview two differently structured portfolios and test whether rigid categories obscure equivalent experience.
Media scheduling and publishing plannerDraft a schedule from licensed sample assets and flag a rights conflict.Week 3 · Expert systems to ChatGPTFogg, Persuasive Technology (2003)LO2 Ethical and societal evaluation; LO4 Applied ethical reasoning; LO7 Communication across audiencesCompare how ranking and interface defaults steer a user's choice of what to publish.
Message reply and personal assistantGenerate and export a reply after human review.Week 3 · Expert systems to ChatGPTFogg, Persuasive Technology (2003)LO1 AI foundations and societal impact; LO4 Applied ethical reasoningCompare suggested drafts and test whether the interface encourages accepting language without review.
Multimodal perception and decision supportCompare results under changed lighting with uncertainty and human override.Week 2 · Cybernetics and grounded cognitionBarsalou, Grounded Cognition (2008)LO1 AI foundations and societal impact; LO3 Case analysis and solutions; LO8 Independent researchCompare image decisions across changes in lighting and context, then discuss what the input fails to ground.
Architecture comparison and graph mappingCompare three system designs against the same requirements.Week 3 · Expert systems to ChatGPTMinsky, Steps Toward Artificial Intelligence (1961) (cross-week application)LO1 AI foundations and societal impact; LO6 Critical discussion; LO7 Communication across audiencesContrast explicit symbolic rules with probabilistic inference and locate where each system represents knowledge.
Model evaluation, bias and reliabilityMeasure disagreement, subgroup errors, latency and cost on synthetic cases.Week 6 · Feedback and model collapseWyllie, Shumailov & Papernot, Fairness Feedback Loops (2024)LO2 Ethical and societal evaluation; LO3 Case analysis and solutions; LO8 Independent researchSimulate repeated selection and measure whether feedback widens an initial subgroup performance gap.
Opportunity window and expenditure calculatorShow how changing assumptions alters rankings and costs.Week 4 · Probability and predictionRescher, Predicting the Future (1998), chs. 7–8LO1 AI foundations and societal impact; LO6 Critical discussion; LO7 Communication across audiencesShow how assumptions change forecasts and compare the conditions under which each forecast holds.
Ga community development metricsTrace one aggregate measure to a dated source.Week 5 · Data as infrastructureCrawford, Atlas of AI (2021), Data chapterLO2 Ethical and societal evaluation; LO7 Communication across audiences; LO9 Interdisciplinary collaborationTrace who defines community categories and report which groups an aggregate measure may hide.
Family history evidence explorerCompare two hypothetical lineage claims with source strength.Week 5 · Data as infrastructureCrawford, Atlas of AI (2021), Data chapterLO4 Applied ethical reasoning; LO7 Communication across audiences; LO8 Independent researchCompare source coverage and conflicting lineage claims while controlling which personal details are exposed.
Distinct Histories introduction appTest consent, disclosure and accessible onboarding with fictional profiles.Week 8 · Fairness and counter-imaginariesKazansky & Milan, Bodies not templates (2021)LO2 Ethical and societal evaluation; LO4 Applied ethical reasoning; LO7 Communication across audiencesTest whether rigid identity categories misrepresent fictional profiles and evaluate a self-description option.
Intentional dating introduction briefGenerate a fictional brief and test data minimisation.Week 3 · Expert systems to ChatGPTFogg, Persuasive Technology (2003)LO2 Ethical and societal evaluation; LO4 Applied ethical reasoning; LO7 Communication across audiencesCompare onboarding defaults and identify where the interface steers disclosure beyond user intent.
Creative portfolio and interactive globeTest keyboard navigation, reduced motion and mobile performance.Week 3 · Expert systems to ChatGPTFogg, Persuasive Technology (2003)LO3 Case analysis and solutions; LO7 Communication across audiencesCompare motion and layout choices for their influence on attention and navigation.
Botanical collection and product catalogueCompare accessible product discovery and mock checkout.Week 3 · Expert systems to ChatGPTFogg, Persuasive Technology (2003)LO3 Case analysis and solutions; LO7 Communication across audiencesCompare catalogue defaults and test whether the ordering steers choices while remaining accessible.
Research manuscript template libraryDownload and open a versioned sample template with provenance.Week 5 · Data as infrastructureMuldoon et al., A typology of artificial intelligence data work (2024) (cross-week application)LO7 Communication across audiences; LO8 Independent researchPreserve version and author information for templates and document the human work of reviewing revisions.
Job application and career supportRecord and export a fictional application and its follow-up.Week 7 · AI, power and the stateBenjamin, Race After Technology (2019), ch. 3LO3 Case analysis and solutions; LO7 Communication across audiencesUse fictional cases to inspect whether recruitment labels could encode proxies for protected characteristics.
Equitable living and public initiatives directoryFind an initiative by geography and verify its source.Week 8 · Fairness and counter-imaginariesKazansky & Milan, Bodies not templates (2021)LO2 Ethical and societal evaluation; LO5 Policy and regulation; LO7 Communication across audiencesTest whether the initiative taxonomy makes communities discoverable on their own terms.
Financial reconciliation and payments sandboxReconcile duplicate and late payment events using synthetic transactions.Week 2 · Cybernetics and grounded cognitionWiener, Cybernetics (1948)LO1 AI foundations and societal impact; LO3 Case analysis and solutions; LO4 Applied ethical reasoningTrace the feedback that detects a repeated payment event and corrects the reconciliation state.
Legacy connector and data migration labRe-run an import without duplicating records and produce an exception report.Week 5 · Data as infrastructureMuldoon et al., A typology of artificial intelligence data work (2024)LO1 AI foundations and societal impact; LO3 Case analysis and solutionsDocument human decisions needed to map old fields and classify ambiguous import records.
Technical demonstration portalCheck source, limitations and test instructions for each demo.Week 9 · Design and accountabilityWinner, Do Artifacts Have Politics? (1980) (cross-week application)LO7 Communication across audiences; LO9 Interdisciplinary collaborationShow how the portal makes test evidence, limitations and responsibility visible across different demos.

Project register

Start with independent chat, automation, CRM, ERP and compliance. Each exercises a different architecture. The linked source pages document the work so far. Starter and test columns describe proposed preparation.

ProjectManifold Grace sourceStarter or architectureFirst acceptance testMA outcomes
Independent ChatGPT-style interfaceSource pageWebLLM minimal chat exampleRun one multi-turn chat, export it, and test cancellation and unsupported devices1, 7
Business process automationSource pageNode-RED flow with synthetic eventsRoute an enquiry through approval and replay a duplicate event1, 3, 4
CRM and Salesforce-style pipelineSource pageFrappe CRM sandboxMove a synthetic lead from enquiry to opportunity with role checks1, 2, 3
ERP and resource planningSource pageERPNext sandboxTrace a test order through stock, invoice and reconciliation1, 3
Compliance and continuous AI assuranceSource pageDjango and SQLite control registerLink a control to evidence, a failed test, remediation and sign-off2, 4, 5
Evidence verification and contradiction APISource pageFastAPI and versioned evidence fixturesBenchmark claim coverage, contradiction and justified abstention1, 3, 8
Research discovery and literature-gap auditSource pagePython search pipeline and SQLiteCompare retrieval against a fixed literature set and test one proposed gap6, 8
Cross-conversation context and private memorySource pageLocal database and retrieval adapterFind a consented record, apply a correction and verify deletion1, 4, 5
Historical as-of retrieval and correction propagationSource pageVersioned relational records and APIReconstruct a claim at two dates and propagate a superseding record3, 5, 8
Document-drop classifier and evidence organiserSource pagePython ingestion, local metadata and file previewClassify mock files and preview filenames before applying any change1, 3
Contributor provenance and innovation attributionSource pageRelational contribution registerRecord origin, revisions and an attribution dispute4, 5, 9
Agency.ai agent workspaceSource pageWorkflow engine and explicit approval queueRun a simulated service request with a reviewable action ledger1, 4, 5
Public-interest research and societal resilience adviserSource pageRetrieval and scenario dashboardCompare two policy scenarios and cite assumptions and evidence gaps2, 6, 7, 8
UK public bodies and oversight directorySource pageStatic directory plus versioned JSON graphTrace a body's remit and oversight across a dated reorganisation5, 7, 9
Learning outcome, viva and competence verificationSource pageRubric and interactive quiz appCompare rubric results on matched sample responses and collect reviewer agreement2, 3, 8, 9
Prior learning and skills evidence portfolioSource pageEvidence-to-outcome matrix and review queueMap one sample portfolio to criteria with a human review decision4, 7, 9
Media scheduling and publishing plannerSource pageCalendar and rights metadataDraft a schedule from licensed sample assets and flag a rights conflict2, 4, 7
Message reply and personal assistantSource pageDraft queue and mock connectorsGenerate and export a reply after human review1, 4
Multimodal perception and decision supportSource pageImage inference and review dashboardCompare results under changed lighting with uncertainty and human override1, 3, 8
Architecture comparison and graph mappingSource pageStatic diagrams and dependency graphCompare three system designs against the same requirements1, 6, 7
Model evaluation, bias and reliabilitySource pagePython benchmark runner and reporting UIMeasure disagreement, subgroup errors, latency and cost on synthetic cases2, 3, 8
Opportunity window and expenditure calculatorSource pageBrowser calculator with test scenariosShow how changing assumptions alters rankings and costs1, 6, 7
Ga community development metricsSource pageStatic dashboard and aggregate datasetTrace one aggregate measure to a dated source2, 7, 9
Family history evidence explorerSource pageVersioned graph and privacy-filtered viewsCompare two hypothetical lineage claims with source strength4, 7, 8
Distinct Histories introduction appSource pageStatic interactive mock-upTest consent, disclosure and accessible onboarding with fictional profiles2, 4, 7
Intentional dating introduction briefSource pageLocal form and exportGenerate a fictional brief and test data minimisation2, 4, 7
Creative portfolio and interactive globeSource pageStatic media portfolio and WebGL prototypeTest keyboard navigation, reduced motion and mobile performance3, 7
Botanical collection and product catalogueSource pageStatic catalogue and simulated basketCompare accessible product discovery and mock checkout3, 7
Research manuscript template librarySource pageStatic download catalogueDownload and open a versioned sample template with provenance7, 8
Job application and career supportSource pageLocal CRUD tracker and document exportRecord and export a fictional application and its follow-up3, 7
Equitable living and public initiatives directorySource pageStatic taxonomy and source registerFind an initiative by geography and verify its source2, 5, 7
Financial reconciliation and payments sandboxSource pageMock webhook listener and ledgerReconcile duplicate and late payment events using synthetic transactions1, 3, 4
Legacy connector and data migration labSource pageCSV fixtures and a Python import adapterRe-run an import without duplicating records and produce an exception report1, 3
Technical demonstration portalSource pageStatic catalogue with runnable demosCheck source, limitations and test instructions for each demo7, 9

2027 venture opportunities

All 28 opportunities from the venture portfolio. These are candidate offerings. Several can use the same project baseline.

OpportunityPotential offering
01. AI evidence verification APIB2B API for claim-to-source coverage, contradictions and dated provenance
02. AI system architecture mappingDiagram-to-dependency maps with implementation discrepancy reviews
03. Agent identity and permission auditsInventory agents, credentials and authorisation boundaries
04. Coding-agent quality assuranceTests, regression monitoring and change provenance for generated code
05. SME workflow implementation studioIndustry-specific automation with training, rollback and support
06. AI procurement due-diligence reportsEvidence-based supplier comparison, data handling and exit-risk dossiers
07. AI context and knowledge governanceVersioned organisational knowledge, retrieval access and correction histories
08. Model and agent expenditure observabilityDashboards for model costs, latency and workflow value
09. AI assurance documentation serviceHuman-reviewed risk registers, evaluation evidence and audit packets
10. AI literacy and verification trainingApplied training in checking outputs and safe workplace integration
11. Accessible AI interface auditingTest accessibility and usability of AI-supported public or business services
12. Human oversight workflow toolsApproval queues, escalation, sign-off and traceability for AI actions
13. Secure AI adoption for charitiesLow-cost setup, staff guidance and data-safe workflows
14. AI cybersecurity evidence triageDefensive summarisation and prioritisation of authorised alert data
15. Provenance and authenticity workflowContent origin, revisions, signatures and source verification
16. Research literature and gap audit assistantFalsification-led evidence maps, overlap detection and reproducible reviews
17. AI-enabled public-systems directoryVersioned public-body relationships, functions and oversight routes
18. AI learning-outcome verificationAssess student understanding rather than AI-generated assignment polish
19. Physical AI simulation evaluationTesting robotics and simulated workflows without selling generic coding toys
20. 2027 opportunity window intelligenceMonitor adoption, unmet need, competitor substitution and window-closing signals
21. Synthetic data & edge-case generationValidated rare-scenario datasets for model and robotics evaluation
22. Legacy-to-AI bridging & middlewarePermissioned data connectors and translation layers for older enterprise systems
23. AI liability and risk analyticsIncident evidence, control assessment and risk telemetry for insurers
24. Human-in-the-loop escalation routingUncertainty-triggered human handoffs, contextual review and approval records
25. On-premise & sovereign AI deploymentLocal or residency-controlled AI stacks with documented security boundaries
26. AI-native application consolidationSubscription inventory, workflow consolidation and value-based software rationalisation
27. Machine-readable brand & commerce optimisationStructured product data and verified metadata compatible with agentic shopping workflows
28. Post-automation workforce reskillingEvidence-based scenarios for human review, system judgement and exception handling

Learning and preparation

The 2024/25 programme specification lists nine outcomes. The numbers above refer to AI foundations (1), ethical evaluation (2), case studies (3), applied ethical reasoning (4), policy and regulation (5), critical discussion (6), communication (7), independent research (8), and interdisciplinary collaboration (9). Outcome 9 needs actual participation and documented feedback.

For each build, record the upstream source and version, licence and notices, installation steps, synthetic fixtures, a working baseline commit, and an experiment branch. Add a data-flow diagram, an acceptance test, a note on what broke, and a record of what you changed. Capture the test result before linking a working demo.

Use WebLLM for browser inference and GitHub Pages for static demos. Run Node-RED, Frappe CRM, ERPNext and Django locally for experiments that need a backend. Verify the specific release and licence before distributing a modified baseline.

Test expected challenges early: replacing a model or provider, recovering from failed workflows, retaining evidence versions, enforcing access rules, exporting data, and keeping the interface accessible.