Konnaxion – Technical Architecture & Services

This page collects the technical details of Konnaxion: service code‑names, core models, configuration parameters, routing invariants, and cross‑module infrastructure. It complements:

• the repository README, which focuses on vision, mythology, and high‑level modules
• the wiki hub page, which explains civic workflows and navigation
• the individual module pages (Knowledge, CertifiKation, Korum, etc.), which describe each sub‑module functionally

Use this page as the reference for development and integration work.

---

1. Platform overview

1.1 KOA module map

Konnaxion’s architecture is organized into six top‑level modules:

• KonnectED – learning, knowledge, and certification
• Ethikos – structured debates and civic consultations
• Kreative – culture, preservation, and professional networks
• keenKonnect – project workspaces and storage
• Kollective Intelligence – reputation and voting (EkoH, Smart Vote)
• System / Core – cross‑cutting auth, storage, search, analytics

Each KOA module is implemented as one or more Django apps with:

• service code‑names (e.g. multidimensional_scoring, public_consultation) mapping 1‑to‑1 to service modules
• OLTP models described in the v14 schema reference
• frozen configuration parameters (thresholds, limits, routes)

1.2 Shared technology stack

Across modules, the technical stack is consistent:

• Backend: Django + Django REST Framework
• Realtime: Django Channels with Redis (channels_redis.core.RedisChannelLayer)
• Data store: PostgreSQL with tsvector full‑text search
• Background tasks: Celery (ETL, AI enrichment, packaging, recomputation)
• Object storage: /app/media/ root, typically backed by S3/MinIO
• Front‑end: routes reserved per module (/learn, /certs, /ethikos/korum, etc.)

---

2. Cross‑module infrastructure

2.1 Service code‑name convention

Every sub‑module defines named services that are stable integration points. Examples:

• Korum: structured_debate, ai_clone_management, comparative_argument_analysis, public_debate_archive, automated_debate_summary
• Smart Vote: dynamic_weighted_vote, voting_modalities, emerging_expert_detection, vote_transparency, vote_result_visualization, cross_module_vote_integration
• EkoH: multidimensional_scoring, configuration_weights, contextual_analysis, privacy_settings, score_history, score_visualization, expertise_field_classification
• Knowledge: library_resource_management, personalized_recommendation, content_co_creation, thematic_forum, learning_progress_tracking

Code‑names map 1‑to‑1 to service modules (e.g. services/dynamic_weighted_vote.py), and are referenced by tasks, API endpoints, and configuration.

2.2 Routing invariants

Top‑level routes are owned by specific modules and treated as invariants. Namespacing is enforced for consistency:

• /learn, /course/[slug] → KonnectED / Knowledge
• /certs → KonnectED / CertifiKation
• /ethikos/korum, /ethikos/insights → Ethikos / Korum
• /ethikos/consult → Ethikos / Konsultations
• /projects, /projects/[slug] → keenKonnect / Konstruct + Stockage
• /kreative, /art/[id], /archive → Kreative / Konservation
• /connect, /profile/[user] → Kreative / Kontact
• /kollective/konsensus, /reports/smart-vote → Kollective Intelligence / Smart Vote

No other module should claim these top‑level paths.

2.3 Storage and media

• Media root: MEDIA_ROOT=/app/media/ is shared across modules.

• File size & types are fixed per context:

  • Stockage / Konstruct: MAX_BLUEPRINT_UPLOAD_MB = 150, types [".pdf", ".png", ".jpg", ".glb", ".gltf", ".stl"]
  • Konservation: ARTWORK_MAX_IMAGE_MB = 50, ARTWORK_RESOLUTIONS = [256, 1024, 2048] px

Storage is typically an object store (S3/MinIO) behind the media path.

2.4 Search

• Knowledge and Stockage explicitly use PostgreSQL full‑text search (SEARCH_BACKEND = "postgres") for library resources and documents.
• Tags (Tag, ArtworkTag) provide a shared taxonomy layer across Kreative and keenKonnect.

2.5 Realtime and background jobs

• Django Channels + Redis power:

  • live stance and result updates in Korum and Konsultations
  • real‑time tallies in Smart Vote
  • project chat, notifications, and document events in Konstruct and Stockage
  • collaboration rooms in Kontact

• Celery tasks & schedules include:

  • etl_smart_vote every 10 minutes, with ~5‑year retention for facts
  • periodic EkoH score recomputation and leaderboard refresh
  • weekly offline packaging for Knowledge (OFFLINE_PACKAGE_CRON = 0 3 * * SUN)
  • image rendition, AI enrichment, and partner ingest for Konservation

---

3. Module–by–module technical summary

3.1 KonnectED

3.1.1 Knowledge – Collaborative Learning Library

Services

• library_resource_management – CRUD and classify KnowledgeResource; enforce type enum.
• personalized_recommendation – compute KnowledgeRecommendation per user.
• content_co_creation – manage CoCreationProject and CoCreationContribution drafts.
• thematic_forum – forums via ForumTopic and ForumPost.
• learning_progress_tracking – track LearningProgress (per user + resource).

Core models

• KnowledgeResource(id, title, type, url, author)
• KnowledgeRecommendation(user, resource, recommended_at)
• LearningProgress(user, resource, progress_percent)
• CoCreationProject, CoCreationContribution
• ForumTopic, ForumPost

Key configuration

• Content types: article | video | lesson | quiz | dataset
• MAX_CONTRIBUTION_DRAFTS = 10 per user
• Search backend: SEARCH_BACKEND = "postgres"
• Offline packaging: OFFLINE_PACKAGE_CRON = 0 3 * * SUN

Routes

• /learn – catalog, recommendations, offline download
• /course/[slug] – course player (lessons, progression)

3.1.2 CertifiKation – Skills & Certification

Services

• certification_path_management – manage CertificationPath.
• automated_evaluation – create Evaluation with raw_score and metadata.
• peer_validation – handle PeerValidation decisions.
• skills_portfolio – connect to Portfolio and core Certificate records.
• certification_interoperability – manage InteropMapping with external LMS/registries.

Core models

• CertificationPath(id, name, description)
• Evaluation(user, path, raw_score, metadata)
• PeerValidation(evaluation, peer, decision)
• Portfolio(user, title, description, items)
• InteropMapping(local_certification, external_system, external_id)
• Certificate (core model for issued credentials)

Key configuration

• CERT_PASS_PERCENT = 80
• QUIZ_RETRY_COOLDOWN_MIN = 30 minutes
• Routes reserved: /certs (Programs, My Certificates)

---

3.2 Ethikos

3.2.1 Korum – Structured Debates

Services

• structured_debate, ai_clone_management, comparative_argument_analysis, public_debate_archive, automated_debate_summary.

Core models

• EthikosCategory – thematic categories
• EthikosTopic – debate topic/question
• EthikosStance(topic, user, value −3…+3)
• EthikosArgument(topic, author, content, parent, side)

Key configuration

• Stance scale: −3 … +3 (0 = neutral)
• Expert cohort quorum: 12 distinct experts (EkoH threshold)
• Moderation auto‑hide: 3 independent reports

Routes

• /ethikos/korum – Korum Hub (Open / Archived / Start New)
• /ethikos/insights – opinion analytics dashboards

3.2.2 Konsultations – Public Consultations & Feedback

Services

• public_consultation, citizen_suggestion, weighted_consultation_vote, consultation_result_visualization, impact_tracking.

Core models

• Consultation(id, title, open_date, close_date, status)
• CitizenSuggestion(consultation, author, content)
• ConsultationVote(user, consultation, raw_value, weighted_value)
• ConsultationResult(consultation, results_data JSONB)
• ImpactTrack(consultation, action, status, date)

Key configuration

• Ballot modalities: approval | ranking | rating | preferential
• Consensus threshold (platform‑wide): ≥ 75% weighted agreement
• Route namespace: /ethikos/consult owned exclusively by Ethikos

Routes

• /ethikos/consult – Consultation Hub (Live / Results / Suggest)
• /ethikos/insights – shared with Korum analytics

---

3.3 Kreative

3.3.1 Konservation – Creative Content & Cultural Preservation

Services

• digital_archive_management, virtual_exhibition, archive_documentation, ai_enriched_catalogue, cultural_partner_integration.

Core models

• KreativeArtwork(id, artist, title, description, media_file, media_type, year, medium, style)
• Tag, ArtworkTag (global tagging vocabulary and join table)
• Gallery, GalleryArtwork(order)
• TraditionEntry(title, description, region, media_file, approved, approved_by)

Key configuration

• ARTWORK_MAX_IMAGE_MB = 50
• ARTWORK_RESOLUTIONS = [256, 1024, 2048]
• VIRTUAL_GALLERY_CAPACITY = 24 artworks/room
• NSFW_FLAG_REQUIRED (bool, default False)
• MEDIA_ROOT = /app/media/

Routes

• /kreative – Creativity Hub
• /art/[id] – Artwork sheet
• /archive – Konservation archive/partners

3.3.2 Kontact – Collaboration & Networking

Services

• professional_profile, intelligent_matching, collaboration_workspace, opportunity_announcement, partner_recommendation.

Core models

• CollabSession(id, name, host, session_type, started_at, ended_at, final_artwork)
• Re‑uses KreativeArtwork, Tag, ArtworkTag for portfolios and tagging.

Key configuration

• COLLAB_CANVAS_MAX_USERS = 6
• MEDIA_ROOT = /app/media/ (shared)
• Routes reserved: /connect, /profile/[user]

Routes

• /connect – People, Opportunities, Collaboration Workspace
• /profile/[user] – public profile/portfolio

---

3.4 keenKonnect

3.4.1 Konstruct – Project Collaboration Spaces

Services

• collaboration_space, project_task_management, real_time_document_editing, integrated_communication, ai_collaboration_analysis.

Core models

• Project(id, title, description, creator, category, status)
• ProjectResource(project, title, url, added_by)
• ProjectTask(project, title, description, assignee, status, due_date)
• ProjectMessage(project, sender, content)
• ProjectTeam(project, user, role, joined_at)
• ProjectRating(project, user, rating, comment)
• Tag (shared)

Key configuration

• MAX_BLUEPRINT_UPLOAD_MB = 150
• ALLOWED_BLUEPRINT_TYPES = [".pdf", ".png", ".jpg", ".glb", ".gltf", ".stl"]
• COLLAB_SPACE_MEMBER_CAP = 40
• AI_SUGGESTION_TOP_N = 8
• VIDEO_SESSION_PROVIDER = "livekit" via KC_VIDEO_PROVIDER env var

Routes

• /projects – Project Studio (Browse, Create, My Projects)
• /projects/[slug] – workspace (Overview, Tasks, Blueprints, Chat, AI Insights, Settings)

3.4.2 Stockage – Secure Repository & Versioned Storage

Services

• secure_document_storage, document_versioning, intelligent_indexing, real_time_sync, granular_permissions.

Core models

• ProjectResource (same as above; canonical file record)
• Project and ProjectTeam for scoping and access control
• Tag for classification

Key configuration

• MAX_BLUEPRINT_UPLOAD_MB = 150
• Allowed types: [".pdf", ".png", ".jpg", ".glb", ".gltf", ".stl"]
• SEARCH_BACKEND = "postgres"
• Realtime layer: channels_redis.core.RedisChannelLayer
• MEDIA_ROOT = /app/media/

Routes

• Exposed inside /projects/[slug] as the “Blueprints” tab.

---

3.5 Kollective Intelligence

3.5.1 EkoH – Reputation & Expertise

Services

• multidimensional_scoring, configuration_weights, contextual_analysis, privacy_settings, score_history, score_visualization, expertise_field_classification.

Core models

• ExpertiseCategory(id, name)
• UserExpertiseScore(user, category, raw_score, weighted_score)
• UserEthicsScore(user, ethical_score)
• ScoreConfiguration(weight_name, weight_value, field)
• ContextAnalysisLog(entity_type, entity_id, field, input_metadata, adjustments_applied)
• ConfidentialitySetting(user, level)
• ScoreHistory(merit_score, old_value, new_value, change_reason)

Key configuration

• Axis weights: quality=1.000, expertise=1.500, frequency=0.750
• Ethical multiplier bounds: floor 0.20, cap 1.50
• EXPERTISE_DOMAIN_CHOICES: 26 ISO‑based domains

Runtime

• Periodic recomputation via Celery Beat; optional realtime pushes of score/leaderboard deltas over Channels+Redis.

3.5.2 Smart Vote – Weighted Voting System

Services

• dynamic_weighted_vote, voting_modalities, emerging_expert_detection, vote_transparency, vote_result_visualization, cross_module_vote_integration.

Core models

• Vote(user, target_type, target_id, raw_value, weighted_value)
• VoteModality(name, parameters JSON)
• EmergingExpert(user, detection_date, score_delta)
• VoteResult(target_type, target_id, sum_weighted_value, vote_count)
• IntegrationMapping(module_name, context_type, mapping_details)

Key configuration

• Modalities: approval | ranking | rating | preferential
• Emerging expert threshold: +15% EkoH delta over 30 days
• Strong consensus threshold: ≥ 75% weighted agreement

Runtime & analytics

• Realtime results through Channels+Redis
• ETL etl_smart_vote every 10 minutes → smart_vote_fact table, 5‑year retention
• UI: /kollective/konsensus (live polls/results), /reports/smart-vote (analytics)

---

4. Data flows and integration

4.1 Reputation‑weighted voting

• EkoH computes per‑user, per‑domain expertise and ethics scores with configurable weights and bounds.

• Smart Vote reads those scores to weight Vote records via dynamic_weighted_vote, adjusting tallies per modality.

• Korum and Konsultations integrate with Smart Vote to obtain EkoH‑weighted stances and ballots:

  • Korum aggregates EthikosStance using EkoH to compute expert cohort views.
  • Konsultations uses weighted_consultation_vote to store raw + weighted values per ballot.

4.2 Projects and documents

• Konstruct manages projects, tasks, chat, and ratings via Project* models.
• Stockage attaches documents and blueprints as ProjectResource records and handles versioning, indexing, and sync.
• Real‑time events (file added/updated/removed; chat messages) are emitted via Channels+Redis to subscribed project workspaces.

4.3 Culture, archives, and networks

• Konservation’s KreativeArtwork, Gallery, and TraditionEntry store creative and heritage outputs with tag‑based discovery.
• Kontact reuses those artefacts and tags for profiles and matching, and stores collaboration sessions in CollabSession.
• AI enrichment and partner ingest tasks update the archive and related metadata in the background.

4.4 Learning and certification

• Knowledge hosts resources, forums, and co‑creation spaces and tracks progression per user/resource.
• CertifiKation uses CertificationPath, Evaluation, and PeerValidation to issue Certificate records and fill user portfolios.
• Although not strictly specified, these activities can feed EkoH via multidimensional_scoring as part of the platform‑wide reputation engine.

---

5. Analytics and insights

• Smart Vote ETL (etl_smart_vote) is the central pipeline for decision analytics, aggregating delta changes from OLTP into a fact table with 5‑year retention, powering /reports/smart-vote.
• Ethikos exposes /ethikos/insights to visualize debate stances and consultation outcomes, consuming Smart Vote facts and Korum/konsultations data.
• EkoH retains a full audit trail via ScoreHistory and ContextAnalysisLog, enabling longitudinal analysis of reputation evolution.

---

6. Contribution guidelines and invariants (technical)

When extending or integrating with Konnaxion, the following invariants should be respected (all documented above):

1. Do not change top‑level route ownership (/learn, /certs, /ethikos/korum, /ethikos/consult, /projects, /kreative, /connect, /kollective/konsensus, /reports/smart-vote) without updating all dependent modules.
2. Preserve service code‑names (e.g. dynamic_weighted_vote, multidimensional_scoring); treat them as public, versioned integration points.
3. Respect frozen parameter values when relying on thresholds, caps, or schedule timings, or introduce new configuration entries in a documented way.
4. Reuse shared infrastructure (Channels+Redis, Celery, /app/media/, PostgreSQL tsvector) to keep behavior consistent and predictable.

This page, together with the module‑specific wiki entries, should provide enough technical context to navigate, extend, and integrate the Konnaxion codebase.