DBIZ Autonomous — The self-operating AI brain for every DBIZ product

DBIZ Autonomous · Self-operating AI Agents

A self-operating AI brain, embedded in every DBIZ product.

Autonomous is a system Shared AI Agents inside ERP, CRM, HCM, MES, TMS: you ask in Vietnamese, and AI understands, queries data, and creates reports, charts, and documents — answers with citations, picks the right model, runs on-premises or in the cloud.

97+ AI Agents
4 capability groups
On-premise or cloud
Reply with citations
Security & international-standards compliance are in the DNA of every DBIZ product
0+
AI Agents, scale on demand
0
capability groups: Core · Dashboard · Enterprise · Personal
0
a knowledge layer, answers with citations
On-prem + Cloud
flexible per your data policy
What is Autonomous

One shared AI brain — not scattered chatbots

Instead of a separate assistant for each app, DBIZ places a shared AI Agents layer beneath the entire ecosystem. Every product calls the same set of capabilities: understanding natural language, reading documents, querying data, generating reports, and producing digital products — consistently and under control.

The AI layer is decoupled from the business logic, so it can scale by GPU, swap/add models and control costs independently — without touching the running system.

Understand & actAsk in Vietnamese → AI queries, analyzes, and answers with the data.
Embedded in every productShared across ERP · CRM · HCM · MES · TMS, not tied to a single app.
Decoupled, scales with GPUsUpgrade to AI without disrupting operations.
How a request gets handled

A standard handling process for every query

Each request goes through 6 clear steps — controlled, traced, and context-saved so the next time is smarter.

1
Load contextpulling history & related data
2
Intent classificationwhat you’re looking for
3
Entity extractiontime, audience, metrics
4
Agent orchestrationpicks the right agent to handle it
5
Generate resultsanswer & generate artifacts
6
Save contextremembers for next time

Orchestrated by LangGraph — Autonomous’s internal processing flow, lightweight and scalable.

4 capability groups

AI Agent groups, organized by what they do

Agents are grouped by capability and usage segment — their number scales flexibly to real-world needs.

Core

Foundational capabilities

Conversation and file processing — the foundation for every other agent.

ConversationRead file · OCRGenerate imageGenerate videoGenerate HTML
Dashboard

Data & reporting

Ask in words — get figures and visual charts instantly.

Text-To-SQLText-To-ChartAutomated reports
Enterprise

Business Agent

Deep-domain agents for each product, understanding enterprise workflows.

ERPCRMMESHCM · TMS…
Personal

Personal assistant

A companion for every employee: reminders, summaries, daily support.

AssistantSummaryTask reminders

The number of agents is deployment variables — add new agents by use case without breaking the architecture.

Not just answering

AI even creates digital products for you

From a single request, Autonomous produces exactly the kind of output you need — then saves it for reuse.

📊
Chartsvisualized from real data
📄
Reportssummarize & analyze
🖼️
Imagesillustrated from a description
🎬
Videoauto-generated short clips
🧩
Trang HTMLtemplates & web assets
P
Platform KBShared knowledge owned by DBIZ — reused across all customers.
T
Tenant KBYour enterprise’s own knowledge — shared across all your agents.
A
Agent KBDedicated knowledge for a specific agent — tied to its task.
3-layer knowledge

AI answers grounded in your own knowledge

Autonomous separates knowledge into 3 clear scopes so agents always use the right source — and every answer is cited sources so you can verify, instead of “making things up”.

🔎 Every answer comes with source citations

Store on the right infrastructure: Oracle 26AI for large enterprises / highly regulated sectors, PostgreSQL for small & medium businesses — with semantic search built in.

Control & Safety

Intelligent, but always in control

Autonomous picks the right model for each task and executes safely — sensitive data never leaves your zone.

🧠 Smart model selection

For each request, the system automatically picks the best model for the context:

Your enterprise’s security policies & levels
Whether there is personal data (PII)
Type of output to generate (text, chart, image, video…)
Balancing cost and quality

Safe execution

Every model call goes through a unified AI serving gateway — with security checks and cost tracking:

Flexible model orchestration on-premise and cloud
A safety filter before data reaches the AI
Trace logging & cost per call
Permissions down to each data row (RLS/CLS)
On-prem modelvLLM · runs in your infrastructure
Model Cloudwhen permitted · for heavy tasks
AI service portalauthentication · guardrails · routing · logging
Agent & Generationget the right model per policy
Guardrails: PII · toxicity · injection SecurityEnforcer: RLS / CLS Cost logging & tracing (Langfuse)
🔒 Offline-first for highly regulated sectors: data and AI traffic run within your zone — never going to the cloud when policy doesn’t allow it.
Self-service · Agent Studio

Businesses configure their own agents

No waiting for reprogramming. With Agent Studio, your team creates and fine-tunes agents on its own — attaching the right tools and knowledge — fully separated from the running operations flow.

Thanks to continuous feedback-driven improvement, the agent grows more accurate the more it’s used.

A
Configure Agentchoose role, scope & behavior
T
Attach tools (Tool)connecting data & the actions needed
K
Attach knowledge (KB)specify the knowledge sources the agent can use
Underlying technology

A modern AI platform, enterprise-ready

Just enough to reassure your IT team — no deep dive needed.

Python · FastAPILangGraphAI Serving Gateway · LiteLLMvLLM (in-house model)Oracle 26AI / PostgreSQLSemantic search (vector)Guardrails & RLS/CLSLangfuse · Prometheus · Grafana

Runs on DBIZ Foundation (Control Plane + Tenant Foundation) — identity, configuration & data per customer.

DBIZ Autonomous

Let AI handle the repetitive work — your people focus on what matters.

Book a 30-minute session: we’ll show you an AI Agent running live on your business’s scenario, and advise your IT team on safe deployment.

Free · no obligation · your data always stays on your infrastructure.