Frequently Asked Questions

Citra AI is a sovereign, privacy-first AI platform for enterprises, built entirely on open-source models. It combines chat, analytics, presentations, reports, dashboards, knowledge graphs, workflow agents, and data integration into one system that runs on-premise or in your private cloud. Your data never leaves your infrastructure and never trains any model.

Both. Citra AI Platform is our licensed software product. Citra Consulting is our end-to-end AI services arm — custom AI agents, chatbot development, model fine-tuning, RAG architecture, AI roadmap design, governance, and MLOps. Most enterprises start with the platform and add consulting for fine-tuning and custom agents.

Yes. Citra Enterprise runs fully on-premise or in your private cloud — air-gapped if required, with zero data egress and a complete audit trail. Citra Cloud is also available as a managed option. Both deliver full data sovereignty.

Open-source models give you zero vendor lock-in, full transparency, predictable cost at scale, and the ability to fine-tune on your domain data. Citra runs the full stack — models, embeddings, vector store, orchestration — entirely in your environment, so your IP never trains a third-party API.

The Data Store — also called Citra Vault — is your secure, AI-indexed knowledge base. Upload PDFs, Word docs, spreadsheets, presentations, images, and audio recordings, and connect SQL databases, MongoDB, S3, and REST APIs. Every chat answer, presentation, report, dashboard, and workflow is grounded in your Data Store with paragraph-level citations.

Seven categories:

• Smart Apps — BA-built line-of-business apps bundling UI + workflow + dashboard, authored in natural language
• Query & Analytics — Enterprise Chat, Quick Chat, Deep Analytics Chat
• Create — Presentation 16:9, Visual Reports, Document Creator (MS Word style), Dashboard
• Visualize Data — Internet Research, Reader & Review, Knowledge Graph, Mindmap, Diagram
• Organize Personal Files — Data Store, Notes, Audio Meeting
• Enterprise Automation & Data Integration — Workflow Agents, Dept Data Flow & MCP, Enterprise Branding
• Tools — Copy Template

A Citra Smart App is a published line-of-business application that bundles three things every team needs: a real UI (forms, queues, detail panels, embedded agent chat), an automated workflow that pre-processes the routine work, and a live KPI dashboard. The whole app is authored by a business analyst in natural language — no developer required — and published at apps.citra.ai/{slug}, wired to your enterprise data via MCP, RAG, and DuckDB. Read the full Smart Apps story.

Low-code platforms still require someone — usually a developer or a heavily trained power user — to lay out screens, model data, wire actions, and build dashboards in a separate BI tool. A Smart App is described in natural language: the BA states the business outcome they want, and the builder agent generates the UI, the workflow, and the KPI dashboard together, coherent with each other and wired to enterprise data. It is not a screen builder; it is an outcome builder.

Yes — by automating routine cases and surfacing only exceptions. Example: a motor-claims Smart App auto-approves claims under $500, runs policy-RAG lookups for high-value claims, and routes them into the BA queue with a recommended decision and a cited policy clause. The BA opens the app once a day to review only items that need human judgment. A team that previously did three hours of daily triage compresses it into a 30-minute morning review. The dashboard tracks SLA risk, monthly payout, and fraud-flag rate live.

Smart Apps include named-approver gates, non-self-approval enforcement, and per-tenant audit logging on every step, every approval, and every decision — built into the platform, not bolted on. Compliance teams get an audit trail by default; security teams get tenant isolation and explicit grant-based sharing; leadership gets the dashboard. The Smart App tier sits on the same governed sovereign stack as the rest of Citra.

Inside your perimeter — on-prem or in your private cloud, exactly like the rest of the Citra platform. Smart Apps connect to your enterprise data through MCP servers, internal SQL, files, lakehouse tables, and RAG indexes. No data leaves your environment, no foreign model trains on your IP, and the published apps.citra.ai/{slug} URL is served by your own deployment.

Three: Enterprise Chat (fast AI chat across your Data Store and the internet), Quick Chat (upload a few files and ask immediately), and Deep Analytics Chat (a sandboxed deep-research analyst that runs multi-step analysis, code execution, and synthesis on your data).

Yes. Upload data to your Data Store and generate full 16:9 presentations, A4 visual reports, Word-style Document Creator output, and interactive drill-down Dashboards in seconds. Every output is grounded in your data with paragraph-level citations. When data changes, one click refreshes everything — no rebuilding, no copy-paste.

Workflow Agents are AI-driven automations that run 24/7 on your infrastructure. Use the visual workflow builder to chain LLM steps, SQL queries, API calls, and approvals — for invoice validation, expense auditing, compliance checks, document generation, and any repeatable knowledge process. Agents read from your Data Store and write back results.

Use Dept Data Flow & MCP to connect SQL (Postgres, MySQL, SQL Server, Oracle), MongoDB, S3 buckets, and any REST API. Each department gets a scoped MCP (Model Context Protocol) server with role-based access, so AI tools can query live enterprise data with full governance.

Yes. Generate knowledge graphs (entity maps), mindmaps, and process diagrams directly from your Data Store. Every node and relationship is extracted from your documents and databases — not invented. Perfect for research synthesis, compliance mapping, and onboarding. See how it works.

Reader & Review is Citra's built-in PDF and document viewer with an AI sidebar. Open any document from your Data Store, highlight passages, and ask the AI questions side-by-side — every answer is grounded in the document with paragraph-level citations.

Yes. Audio Meeting records meetings, transcribes them, and adds them to your Data Store as searchable, citable knowledge. Combine with Notes and uploaded files to build a complete project memory.

End-to-end AI services:

• Custom AI Agents — built for your specific workflows
• Chatbot Development — domain-specific conversational bots
• Model Fine-Tuning — train open-source models on your proprietary data
• RAG Hybrid Architecture & Rollout — retrieval pipelines tuned to your domain
• AI Roadmap Design — strategy and execution across the organization
• MLOps, Governance & Security — production-grade operations

Engagements are flexible — fixed-scope projects or ongoing capacity. Billed at $20 / hour.

Three phases:

• 01 · Pilot (0–3 months) — deploy on your infrastructure, prove ROI on one high-value workflow.
• 02 · Scale (3–9 months) — expand to 3–5 workflows across departments, train domain models, onboard 100+ users.
• 03 · Enterprise Rollout (9+ months) — multi-region deployment, governance, audit, MLOps, AI embedded across the organization.

Generic SaaS AI tools send your data to vendor servers, charge $60–$120 per user per month, and offer no on-prem option. Citra runs in your environment on open-source models, costs $20 per user per month, and grounds every output in your own data with citations. It also covers chat, analytics, dashboards, presentations, workflow agents, and MCP — not just one workflow. See a detailed comparison.

Yes — Citra Cloud Free lets you try chat, presentations, reports, and Data Store with no credit card. For on-premise enterprise deployment, contact us for a live demo and custom scope.

Yes. Citra supports multi-user workspaces, shared Data Stores, role-based access controls, and Enterprise Branding (logo, colors). Each team member has their own workspace while collaborating on shared knowledge and outputs.