Markdown source Marp deck Marketing index

v2.3.0 Local-first AI Enterprise governance

Sovereign LLM Workbench

Local AI for confidential documents — Q&A, reports, and diagrams without sending data to the cloud.

Marketing presentation · June 2026

The problem

PainBusiness impact
Data sovereigntyClient IP and regulated data cannot leave the perimeter
Unpredictable costPer-token API bills grow with every analyst
Hallucination & trustGeneric chat is not grounded in your policies
Vendor lock-inOutages and policy changes block critical work

Teams resort to manual PDF search or risky use of public ChatGPT.

The solution

Your documents → RAG index → Grounded answers → DOCX/PPTX + diagrams → Audit trail ↑ Local LLM (LM Studio via Docker llmster; failover URL)

Feature list

Core

  • Multi-turn chat with conversation store
  • RAG query with hybrid search & source citations
  • Knowledge base — PDF, DOCX, TXT, MD, CSV, XLSX
  • DOCX / PPTX reports · Mermaid / PlantUML diagrams
  • Streaming chat (SSE) · prompt library & attachments

Integrations & government (v2.3+)

  • VS Code / Cursor extension + MCP
  • Chat modes & auto-routing (Agent, Plan, Ask, Auto)
  • Government productivity — multilingual, GR, meeting minutes
  • Sovereign readiness · model routing · agent tools
  • Enhanced RAG — reranking, retrieval guard

Enterprise (v2.3)

  • Portal · OIDC JWT + SAML SP · RBAC · audit chain
  • GRC repository · SLHP scheduler · regulatory KB ingest
  • Security engineering & Red Team APIs
  • LLM failover · multi-tenant · OT templates
  • Vision / PDF OCR · Enterprise ML stack builder

Capabilities list

LayerHighlights
LLMLM Studio (primary); failover URL; streaming; JSON reports; model routing
RAGChromaDB + local embeddings; token-aware chunking; anti-hallucination prompts
Documentspython-docx / python-pptx; diagram CLI render to PNG/SVG
SecurityAPI key, OIDC, RBAC, SSRF-safe scrape, air-gap defaults
GovernanceOWASP LLM, NIST AI RMF, EU AI Act, gap assessment, SLHP scheduler, SBOM/AIBOM

GET /capabilities · GET /governance/assurance/summary

Use cases

IDScenarioOutcome
UC-01Internal policy Q&ASourced answers for audit & ops
UC-02Compliance / gap reportDOCX/PPTX in minutes
UC-03Controlled web ingestExpand KB when policy allows
UC-04Architecture diagramsPNG/SVG deliverables
UC-05General chatDrafting without KB
UC-06GRC evidence packFramework crosswalk + trust index
UC-09Regulatory / law Q&ACited official texts in KB
UC-10Security gap & Red TeamGap API + exercise program

Case studies (representative scenarios)

Illustrative pilot-style scenarios — substitute your customer logos and metrics after deployment.

ScenarioChallengeResult
Consulting firmChatGPT used on client docsZero egress; faster DOCX summaries; audit trail
Financial compliance200+ policy PDFsHours → minutes to answer; explicit “no context”
Air-gapped agencyNo cloud AI allowedFull offline workflow after model install
OT / industrialUngoverned AI tools98% unified assurance; OT deploy templates

Cost optimisation

DriverCloud AISovereign
InferencePer-token / seatFixed hardware + open models
Data egressImplicitNone
IntegrationMultiple toolsSingle workbench + GRC APIs

Illustrative savings (10 analysts, year 1)

Avoid $6k–$24k+ in team cloud-AI subscriptions* · Amortized hardware $2k–$5k

*Illustrative; adjust for your market. See PRESENTATION.md for full TCO table.

Competitive analysis

vsYou win when…
Cloud AI (ChatGPT, Claude)Sovereignty, predictable cost, native RAG + reports matter most
DIY RAG (LangChain)You need productized UI, audit, and GRC in days not months
Suite AI (Copilot, Duet)M365/Workspace lock-in is unacceptable; air-gap required
Vertical compliance SaaSYou need flexible corpus + AI deliverables + framework crosswalk

Frontier cloud models still lead on complex reasoning; Sovereign leads on control and TCO.

Proof points

100%
Production readiness
100%
AI governance
98%
Cyber posture
97
AI Trust Index
98%
Unified assurance
EditionTarget
CommunityPilot / individual
ProfessionalSmall teams · Docker
EnterpriseRegulated · SSO · SLA

5-minute demo & call to action

  1. Ingest demo/sample_policy.txt
  2. RAG: “What is the data retention policy?”
  3. Generate DOCX incident-response summary
  4. Show /health, audit log, /governance/assurance/summary

Next steps