DEV Community: tajdin k

Noviz AI Agent for ERPNext — A Reasoning Layer for ERPNext

tajdin k — Wed, 02 Sep 2026 11:46:58 +0000

# Noviz AI Agent for ERPNext — A Reasoning Layer for ERPNext

Noviz AI is an AI Agent built specifically for ERPNext.

Instead of simply answering questions, Noviz is designed to understand user intent, reason about the ERPNext environment, and work with existing ERPNext capabilities to accomplish tasks.

How It Works

User → Noviz AI → ERPNext

The user provides the intent, Noviz handles the reasoning, and ERPNext remains the underlying source of business data, permissions, DocTypes, workflows, reports, and actions.

What Developers Can Expect

Work with ERPNext DocTypes and records
Analyze ERPNext data using natural-language requests
Generate reports, tables, and charts
Execute actions on ERPNext records
Work with workflows and business processes
Handle notification-driven processes
Coordinate multi-step ERPNext operations
Work with relevant ERP and email context

Built Around ERPNext

Noviz does not attempt to replace ERPNext's existing architecture.

It is designed to work with capabilities ERPNext already provides, including:

DocTypes
Metadata
APIs
User permissions
Reports
Workflows
Notifications
Business data

Permission-Aware

Noviz operates within the permissions of the currently logged-in ERPNext user.

The data and actions available through the agent are governed by the user's existing ERPNext access.

Your ERPNext login. Your access. Your permissions.

Data Efficiency

Noviz is designed so that the complete ERP database does not need to be sent to a language model for every request.

Processing can take place within the ERPNext environment, while AI reasoning can use metadata and only the relevant context required for the task.

From AI Assistant to AI Agent

The goal is not simply to build another chatbot for ERPNext.

Noviz is being developed around a different approach:

Understand the intent → Reason about the ERP → Use ERPNext capabilities → Get the task done.

This makes natural-language interaction a way to work with ERPNext, rather than simply ask questions about it.

Now Available

Noviz AI is now available on the Frappe Marketplace:

https://cloud.frappe.io/marketplace

We would love to hear from Frappe developers, ERPNext implementers, consultants, and businesses about real-world use cases, feedback, and ideas for where an AI Agent can be useful inside ERPNext.

Noviz AI — Your AI Agent for ERPNext.

Building an AI Layer for ERPNext: Using Frappe as the Execution Engine

tajdin k — Tue, 25 Aug 2026 07:47:14 +0000

I have been working on Noviz AI, an AI agent integration for ERPNext, and one architectural principle has become particularly important during development:

The LLM should reason about the user’s request, but it should not become the database, reporting engine, or visualization engine.

ERPNext and Frappe already provide most of the infrastructure required to execute business operations reliably. The AI layer can sit above that infrastructure and use it through a controlled tool interface.

Architecture
The integration is intentionally divided into separate responsibilities.

The LLM / AI layer understands the user’s intent, reasons about the required operations, selects the appropriate tools, determines filters, grouping and metrics, and interprets the returned results.

The Frappe execution layer performs the actual DocType operations, queries, aggregations, pagination and analytics while operating within Frappe’s existing permission model.

ERPNext remains the source of truth for the underlying business data, DocTypes, relationships, permissions, reports and document-generation capabilities.

This separation is important because the LLM is not being turned into an ERP database engine.

simplechart
simplechart
871×784 35 KB
Generating reports without sending the entire dataset to the LLM
One of the areas we have focused on is analytics and reporting.

Suppose a user requests a customer-wise sales analysis.

The reasoning layer can translate the request into an analytical operation containing the relevant entity, filters, grouping field, metric, aggregation and pagination requirements.

The execution layer performs that operation directly against ERPNext.

The complete result does not necessarily need to become part of the LLM’s context. The model can receive a compact representation of the analytical result, while the complete dataset remains available to the application for presentation.

This creates a useful separation:

LLM context → information required for reasoning

UI / report layer → complete information required for presentation

That becomes increasingly important when dealing with hundreds or thousands of records.

Tables, charts and reports as reusable capabilities
Tables, charts and reports do not have to become separate hard-coded AI features.

An analytics operation can produce structured data, and that same result can be consumed by different presentation capabilities.

A table can display the complete dataset. A chart can visualize the aggregated values. A report can use the same underlying query definition, and the report can subsequently be rendered as a PDF.

The LLM’s role is primarily to determine what needs to be calculated and how the result should be presented.

The deterministic execution and presentation layers perform the actual work.

This makes the system much more composable. New reporting or visualization capabilities can be added without requiring the reasoning layer to understand every implementation detail.

Reducing unnecessary token usage
This architecture also has an important effect on LLM costs.

A naive ERP AI implementation could retrieve a large number of records and repeatedly place those records into the model’s context during multi-step reasoning.

That is expensive and often unnecessary.

Instead, Noviz separates:

data retrieval → analytical processing → presentation → reasoning

For example, an aggregated result may contain hundreds of groups, but the reasoning model may only need summary information and pagination metadata to decide what to do next.

The complete dataset can remain outside the model context and be rendered directly to the user.

The LLM can therefore concentrate on the part where it provides the most value: reasoning and decision-making.

Working with Frappe’s existing permission model
Another important aspect is security.

The public Noviz Frappe application does not maintain a separate ERP credential for performing operations on behalf of the user.

The execution request is handled by the Frappe application under the authenticated user’s session. The existing ERPNext permission system therefore remains an important enforcement boundary.

The AI can request an operation, while Frappe determines whether that operation is permitted for the current user.

This keeps authorization close to the system that owns the data rather than attempting to recreate ERPNext’s permission model inside the AI layer.

                 User Request
                      │
                      ▼
            ┌──────────────────┐
            │   LLM / AI Layer │
            │                  │
            │ Intent + Reasoning
            │ Tool Selection   │
            │ Planning         │
            └────────┬─────────┘
                     │
               Tool / Call Spec
                     │
                     ▼
          ┌──────────────────────┐
          │ Frappe Execution     │
          │                      │
          │ DocTypes / ORM       │
          │ Filters / Queries    │
          │ Permissions          │
          │ Analytics            │
          └──────────┬───────────┘
                     │
                 ERPNext Data
                     │
          ┌──────────┼──────────┐
          ▼          ▼          ▼
        Table      Chart      Report
                                 │
                                 ▼
                                PDF

Why this architecture fits Frappe particularly well
Frappe already provides the fundamental building blocks required for this type of AI integration: DocTypes, ORM, permissions, reports, print formats, APIs and the Desk environment.

The AI layer therefore does not need to rebuild an ERP abstraction from scratch.

Instead, it can act as an intelligent orchestration layer over the existing Frappe ecosystem.

A generic execution protocol is particularly useful here. Rather than exposing every internal implementation detail to the model, the AI platform can translate its reasoning into controlled operations such as retrieving documents, retrieving lists, creating or updating documents, and executing analytical operations.

The Frappe application then performs those operations using its native mechanisms.

A useful direction for AI + Frappe
This architecture opens an interesting direction for ERPNext.

Instead of building an AI feature for every individual report, dashboard or analytical question, the AI can be given a well-designed set of composable capabilities.

The intelligence determines the required operation.

Frappe executes it.

The analytics layer processes the data.

The UI renders the result.

The goal is not to replace ERPNext’s reporting or business logic. It is to make those existing capabilities accessible through a reasoning interface.

This makes it possible to build increasingly sophisticated natural-language workflows while keeping the underlying execution deterministic, permission-aware and efficient in terms of LLM context.

For AI integrations with Frappe, the separation between reasoning and execution provides a practical foundation for natural-language reporting, analytics, charts, tables, PDFs and broader ERP workflows

𝗔𝗜 𝗔𝗴𝗲𝗻𝘁𝘀 𝗶𝗻 𝘁𝗵𝗲 𝗘𝗻𝘁𝗲𝗿𝗽𝗿𝗶𝘀𝗲: 𝗕𝘂𝗶𝗹𝗱𝗶𝗻𝗴 𝗦𝗺𝗮𝗿𝘁𝗲𝗿 𝗘𝗥𝗣 𝗗𝗮𝘀𝗵𝗯𝗼𝗮𝗿𝗱𝘀 𝘄𝗶𝘁𝗵 𝗣𝘂𝗿𝗽𝗼𝘀𝗲 𝗕𝘂𝗶𝗹𝘁 𝗧𝗼𝗼𝗹𝘀

tajdin k — Sun, 16 Aug 2026 06:28:53 +0000

🚀

One of the new capabilities we’ve been working on in our ERP AI agent is 𝗔𝗜 𝗴𝗲𝗻𝗲𝗿𝗮𝘁𝗲𝗱 𝗱𝗮𝘀𝗵𝗯𝗼𝗮𝗿𝗱 𝘃𝗶𝗲𝘄𝘀, allowing users to describe the view they need and letting the agent construct it from the underlying ERP data.
The key challenge wasn’t simply getting an LLM to generate a dashboard.
It was making the generation 𝗮𝗰𝗰𝘂𝗿𝗮𝘁𝗲, 𝗰𝗼𝗻𝘀𝗶𝘀𝘁𝗲𝗻𝘁 𝗮𝗻𝗱 𝗲𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝘁.
We addressed this by defining 𝗽𝘂𝗿𝗽𝗼𝘀𝗲 𝗯𝘂𝗶𝗹𝘁, 𝘁𝘆𝗽𝗲𝗱 𝘁𝗼𝗼𝗹 𝗰𝗮𝗹𝗹𝘀 around the actual operations required to build a dashboard.
The LLM orchestrates the workflow, while the tools handle the deterministic work.
🔹 𝗙𝗲𝘁𝗰𝗵
Retrieves and aggregates the required ERP data through operations such as 𝗴𝗿𝗼𝘂𝗽𝗕𝘆, counts, sums, correlations and other supported aggregations.
🔹 𝗦𝗵𝗮𝗽𝗲
Converts those results into a validated chart and view specification. Server side calculations, category limits and structural constraints are handled outside the model.
🔹 𝗥𝗲𝗻𝗱𝗲𝗿
Converts the validated specification into the actual dashboard components and desktop view.
This gives the LLM a more precise responsibility.
Rather than asking the model to retrieve data, calculate metrics, design charts and implement the UI in a single generation, it determines 𝘄𝗵𝗶𝗰𝗵 𝗼𝗽𝗲𝗿𝗮𝘁𝗶𝗼𝗻𝘀 𝗮𝗿𝗲 𝗿𝗲𝗾𝘂𝗶𝗿𝗲𝗱 and orchestrates the corresponding tool calls.
For example:
𝗖𝗿𝗲𝗮𝘁𝗲 𝗮 𝘀𝗮𝗹𝗲𝘀 𝗱𝗮𝘀𝗵𝗯𝗼𝗮𝗿𝗱 𝘀𝗵𝗼𝘄𝗶𝗻𝗴 𝗼𝗿𝗱𝗲𝗿 𝘀𝘁𝗮𝘁𝘂𝘀, 𝘁𝗼𝗽 𝗽𝗿𝗼𝗱𝘂𝗰𝘁𝘀 𝗮𝗻𝗱 𝗺𝗼𝗻𝘁𝗵𝗹𝘆 𝘀𝗮𝗹𝗲𝘀 𝘁𝗿𝗲𝗻𝗱𝘀.
The agent can fetch the required datasets, shape each visualization and compose the resulting components into a single desktop dashboard.
We tested the approach against real dashboard generation, focusing on 𝗮𝗰𝗰𝘂𝗿𝗮𝗰𝘆, 𝗲𝘅𝗲𝗰𝘂𝘁𝗶𝗼𝗻 𝗲𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝗰𝘆 and 𝗰𝗼𝗻𝘀𝗶𝘀𝘁𝗲𝗻𝗰𝘆.
One important finding was that reducing the implementation burden on the LLM can improve the result.
The model focuses on 𝗶𝗻𝘁𝗲𝗻𝘁 𝗮𝗻𝗱 𝗼𝗿𝗰𝗵𝗲𝘀𝘁𝗿𝗮𝘁𝗶𝗼𝗻, while deterministic operations are handled by tools that can be validated and tested independently.
Another important aspect is 𝗧𝗼𝗼𝗹 𝗖𝗼𝘃𝗲𝗿𝗮𝗴𝗲.
In several cases, the model correctly identified the required operation, but the corresponding ERP entity or capability was not exposed through the tool layer.
This is the direction we’re taking with the new dashboard capabilities:
𝗡𝗮𝘁𝘂𝗿𝗮𝗹 𝗹𝗮𝗻𝗴𝘂𝗮𝗴𝗲 𝗶𝗻𝘁𝗲𝗻𝘁 → 𝗽𝗿𝗲𝗰𝗶𝘀𝗲 𝘁𝗼𝗼𝗹 𝗰𝗮𝗹𝗹𝘀 → 𝘃𝗲𝗿𝗶𝗳𝗶𝗲𝗱 𝗱𝗮𝘁𝗮 𝗼𝗽𝗲𝗿𝗮𝘁𝗶𝗼𝗻𝘀 → 𝗰𝗼𝗺𝗽𝗼𝘀𝗲𝗱 𝗲𝗻𝘁𝗲𝗿𝗽𝗿𝗶𝘀𝗲 𝗱𝗮𝘀𝗵𝗯𝗼𝗮𝗿𝗱.
𝗜𝘁 𝗶𝘀 𝗮𝗯𝗼𝘂𝘁 𝗴𝗶𝘃𝗶𝗻𝗴 𝘁𝗵𝗲 𝗟𝗟𝗠 𝘁𝗵𝗲 𝗿𝗶𝗴𝗵𝘁 𝗮𝗯𝘀𝘁𝗿𝗮𝗰𝘁𝗶𝗼𝗻𝘀 𝘁𝗼 𝗱𝗼 𝗶𝘁𝘀 𝗽𝗮𝗿𝘁 𝗽𝗿𝗲𝗰𝗶𝘀𝗲𝗹𝘆.

AI #AIAgents #LLM #EnterpriseAI #ERP #SoftwareEngineering #DataViz #OpenSource

🚀 𝗔𝗻 𝗢𝗽𝗲𝗻-𝗦𝗼𝘂𝗿𝗰𝗲 𝗙𝗼𝘂𝗻𝗱𝗮𝘁𝗶𝗼𝗻 𝗳𝗼𝗿 𝗕𝘂𝗶𝗹𝗱𝗶𝗻𝗴 𝗔𝗜 𝗔𝗴𝗲𝗻𝘁𝘀 𝗳𝗼𝗿 𝗘𝗥𝗣 𝗘𝗥𝗣 𝗔𝗴𝗲𝗻𝘁 𝗔𝗜

tajdin k — Sat, 15 Aug 2026 18:29:11 +0000

🚀 𝗔𝗻 𝗢𝗽𝗲𝗻-𝗦𝗼𝘂𝗿𝗰𝗲 𝗙𝗼𝘂𝗻𝗱𝗮𝘁𝗶𝗼𝗻 𝗳𝗼𝗿 𝗕𝘂𝗶𝗹𝗱𝗶𝗻𝗴 𝗔𝗜 𝗔𝗴𝗲𝗻𝘁𝘀 𝗳𝗼𝗿 𝗘𝗥𝗣
𝗘𝗥𝗣 𝗔𝗴𝗲𝗻𝘁 𝗔𝗜 is now available as a 𝗳𝗿𝗲𝗲 𝗮𝗻𝗱 𝗼𝗽𝗲𝗻-𝘀𝗼𝘂𝗿𝗰𝗲 𝘀𝗼𝘂𝗿𝗰𝗲-𝗰𝗼𝗱𝗲 𝗽𝗿𝗼𝗷𝗲𝗰𝘁 for developers and teams building AI agents for 𝗘𝗥𝗣𝗡𝗲𝘅 and enterprise ERP systems.
🔗 https://lnkd.in/gy8SMnWs
The idea is simple:
𝗗𝗼𝗻’𝘁 𝘀𝘁𝗮𝗿𝘁 𝗯𝘂𝗶𝗹𝗱𝗶𝗻𝗴 𝗮𝗻 𝗘𝗥𝗣 𝗔𝗜 𝗔𝗴𝗲𝗻𝘁 𝗳𝗿𝗼𝗺 𝘀𝗰𝗿𝗮𝘁𝗰𝗵.
Start with a working architecture, understand the source code, and build your own capabilities on top of it.
𝗪𝗵𝗮𝘁’𝘀 𝗶𝗻𝗰𝗹𝘂𝗱𝗲𝗱?
→ 𝗟𝗟𝗠 & 𝗔𝗴𝗲𝗻𝘁 𝗢𝗿𝗰𝗵𝗲𝘀𝘁𝗿𝗮𝘁𝗶𝗼𝗻
→ 𝗧𝗼𝗼𝗹 𝗖𝗮𝗹𝗹𝗶𝗻𝗴 & 𝗘𝘅𝗲𝗰𝘂𝘁𝗶𝗼𝗻
→ 𝗕𝘂𝘀𝗶𝗻𝗲𝘀𝘀 𝗖𝗼𝗻𝘁𝗲𝘅𝘁
→ 𝗥𝗼𝗹𝗲-𝗕𝗮𝘀𝗲𝗱 𝗔𝘂𝘁𝗵𝗼𝗿𝗶𝘇𝗮𝘁𝗶𝗼𝗻
→ 𝗠𝗼𝗱𝘂𝗹𝗲 & 𝗘𝗻𝘁𝗶𝘁𝘆 𝗥𝗲𝗴𝗶𝘀𝘁𝗿𝘆
→ 𝗪𝗼𝗿𝗸𝗳𝗹𝗼𝘄 & 𝗥𝗲𝗽𝗼𝗿𝘁 𝗙𝗮𝗰𝘁𝗼𝗿𝗶𝗲𝘀
→ 𝗘𝗥𝗣 𝗖𝗼𝗻𝗻𝗲𝗰𝘁𝗼𝗿 𝗟𝗮𝘆𝗲𝗿
→ 𝗔𝗴𝗲𝗻𝘁 𝗜𝗻𝘁𝗲𝗿𝗳𝗮𝗰𝗲
→ 𝗔𝗱𝗺𝗶𝗻𝗶𝘀𝘁𝗿𝗮𝘁𝗶𝗼𝗻 & 𝗖𝗼𝗻𝗳𝗶𝗴𝘂𝗿𝗮𝘁𝗶𝗼𝗻
𝗘𝗥𝗣𝗡𝗲𝘅𝘁 𝗥𝗲𝗳𝗲𝗿𝗲𝗻𝗰𝗲 𝗜𝗺𝗽𝗹𝗲𝗺𝗲𝗻𝘁𝗮𝘁𝗶𝗼𝗻
The repository includes a working ERPNext connector and an end-to-end reference workflow.
𝗨𝘀𝗲𝗿 → 𝗔𝗜 𝗔𝗴𝗲𝗻𝘁 → 𝗧𝗼𝗼𝗹 → 𝗘𝗥𝗣𝗡𝗲𝘅𝘁 → 𝗕𝘂𝘀𝗶𝗻𝗲𝘀𝘀 𝗗𝗮𝘁𝗮
The open-source implementation demonstrates the 𝗦𝗮𝗹𝗲𝘀 𝗨𝘀𝗲𝗿 / 𝗤𝘂𝗼𝘁𝗮𝘁𝗶𝗼𝗻 𝗟𝗶𝘀𝘁 capability.
Installation and testing documentation is included so developers can 𝗶𝗻𝘀𝘁𝗮𝗹𝗹, 𝗿𝘂𝗻, 𝘁𝗲𝘀𝘁 𝗮𝗻𝗱 𝗲𝘅𝘁𝗲𝗻𝗱 the project.
This is not just an architecture diagram.
𝗧𝗵𝗲 𝘀𝗼𝘂𝗿𝗰𝗲 𝗰𝗼𝗱𝗲 𝗶𝘀 𝗽𝘂𝗯𝗹𝗶𝗰.
𝗪𝗵𝘆 𝗶𝘀 𝘁𝗵𝗶𝘀 𝘂𝘀𝗲𝗳𝘂𝗹?
If you are building an ERP AI Agent, you can use the project as a starting point for:
🔹 ERPNext AI development
🔹 Custom ERP agents
🔹 LLM tool-calling experiments
🔹 ERP business workflows
🔹 Enterprise AI integrations
🔹 AI-powered ERP applications
🔹 Research and development around AI + ERP
𝗙𝗿𝗲𝗲 & 𝗢𝗽𝗲𝗻 𝗦𝗼𝘂𝗿𝗰𝗲
The project is available under the 𝗔𝗚𝗣𝗟-𝟯.𝟬 license.
𝗚𝗶𝘁𝗛𝘂𝗯:
https://lnkd.in/gy8SMnWs
𝗕𝘂𝗶𝗹𝘁 𝗼𝗻 𝘁𝗵𝗲 𝗦𝗮𝗺𝗲 𝗙𝗼𝘂𝗻𝗱𝗮𝘁𝗶𝗼𝗻 𝗮𝘀 𝗡𝗼𝘃𝗶𝘇
The open-source ERP Agent AI project is part of the technology foundation behind 𝗡𝗼𝘃𝗶𝘇 — our broader ERP AI platform.
🌐 https://noviz.in/
If you are building an AI Agent for ERP, 𝗱𝗼𝗻’𝘁 𝗯𝘂𝗶𝗹𝗱 𝗲𝘃𝗲𝗿𝘆𝘁𝗵𝗶𝗻𝗴 𝗳𝗿𝗼𝗺 𝘇𝗲𝗿𝗼.
𝗦𝘁𝗮𝗿𝘁 𝘄𝗶𝘁𝗵 𝘁𝗵𝗲 𝘀𝗼𝘂𝗿𝗰𝗲.
𝗦𝘁𝘂𝗱𝘆 𝘁𝗵𝗲 𝗮𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝘂𝗿𝗲.
𝗥𝘂𝗻 𝘁𝗵𝗲 𝗘𝗥𝗣𝗡𝗲𝘅𝘁 𝗶𝗺𝗽𝗹𝗲𝗺𝗲𝗻𝘁𝗮𝘁𝗶𝗼𝗻.
𝗕𝘂𝗶𝗹𝗱 𝘆𝗼𝘂𝗿 𝗼𝘄𝗻 𝗘𝗥𝗣 𝗔𝗜 𝗔𝗴𝗲𝗻𝘁.
🔗 𝗢𝗽𝗲𝗻 𝗦𝗼𝘂𝗿𝗰𝗲: https://lnkd.in/gy8SMnWs

AIAgents #ERPNext #ERP #EnterpriseAI #AgenticAI #OpenSource #LLM #AIEngineering #ERPIntegration #SoftwareArchitecture

AI Agent Architecture for ERPNext and Enterprise ERP Systems

tajdin k — Fri, 14 Aug 2026 08:47:52 +0000

I’ve been working on Noviz AI, a developer-focused platform for connecting AI agents with enterprise ERP systems.

The idea behind Noviz is simple:

Instead of building a separate AI integration from scratch for every ERP, provide a reusable AI-agent and ERP connector architecture that developers can extend and customize.

ERPNext is the first reference implementation.

What is Noviz AI?

Noviz is an ERP AI agent architecture and complete ERP connector source-code platform, built initially for ERPNext.

It connects an AI agent with ERP business data, APIs, workflows and operations, allowing users to interact with their ERP through natural language while the agent works with the underlying business system.

The architecture is designed to be extended beyond ERPNext to other ERP platforms.

The long-term vision is to provide a common architecture for building AI agents that can operate across different enterprise systems.

Why Noviz?

ERP systems contain the core operational information of a business:

Customers
Sales orders
Invoices
Products
Inventory
Suppliers
Employees
Accounting
CRM
Projects
Workflows
Reports

But accessing this information normally requires users to navigate through ERP interfaces, reports, filters and multiple screens.

An AI agent can provide a different interface.

Instead of:

Open ERP → Find module → Select filters → Find records → Read report

A user can ask:

"Which customers have overdue invoices?"

or:

"Show me the sales orders that are pending delivery."

or:

"Which products are currently low in stock?"

The agent can understand the request, retrieve the appropriate business context and interact with the ERP through its available tools and APIs.

Noviz architecture

The core Noviz architecture can be viewed as:

                    User
                      │
                      ▼
                AI Agent / LLM
                      │
                      ▼
             Noviz Agent Layer
                      │
          ┌───────────┼───────────┐
          ▼           ▼           ▼
       Context      Tools      Workflows
          │           │           │
          └───────────┼───────────┘
                      ▼
              Noviz ERP Connector
                      │
                      ▼
                   ERPNext
                      │
                      ▼
             Enterprise Data

The important part is that the LLM is not directly connected to the database and allowed to do whatever it wants.

Noviz provides an intermediate architecture where the agent can work through defined context, tools, APIs and business operations.

ERPNext is the reference implementation

I chose ERPNext as the initial ERP implementation because it provides a strong open-source ERP environment with APIs, business modules and structured enterprise data.

Noviz can use ERPNext as the underlying business system while providing the AI-agent layer above it.

This creates a practical reference implementation that demonstrates how an AI agent can work with a real ERP rather than just answering questions about static documents.

You can explore the Noviz implementation and live demonstration at:

https://noviz.in/

From AI chatbot to ERP agent

There is an important difference between an AI chatbot and an ERP AI agent.

A chatbot might answer:

"What is the status of my orders?"

An ERP agent needs to actually understand the ERP context, retrieve the relevant records and potentially perform actions.

The Noviz architecture is therefore built around:

Intent → Context → Tools → ERP → Result

For more complex workflows, this can become:

Understand → Plan → Retrieve → Execute → Verify → Report

That is where I see the real opportunity for enterprise AI agents.

Designed to extend beyond ERPNext

Although ERPNext is the current implementation, Noviz is not intended to be limited to ERPNext.

The connector architecture is designed so that the ERP-specific integration can be replaced or extended.

Conceptually:

                    Noviz AI Agent
                          │
                ┌─────────┴─────────┐
                │                   │
        ERP Connector Layer   Business Context
                │
       ┌────────┼────────┬─────────┐
       ▼        ▼        ▼         ▼
    ERPNext     SAP     Odoo    Custom ERP

This means developers and ERP companies can potentially adapt the Noviz architecture to their own ERP environment.

Complete source code

One of the main goals of Noviz is to make the architecture available as a complete source-code product.

Rather than providing only an API or hosted AI assistant, Noviz is intended for developers, ERP companies, system integrators and organizations that want to own and customize the underlying implementation.

The source-code architecture can be extended for:

Custom ERP integrations
Additional AI models
New ERP tools
Business workflows
Custom context providers
Enterprise authentication
Permission-aware operations
Industry-specific agents
Custom automation

The ERPNext implementation serves as the starting point and reference architecture.

Who is Noviz for?

I see Noviz being useful for several groups:

ERP developers

Developers who want to add AI capabilities to an existing ERP without designing the entire agent architecture from zero.

ERP implementation companies

Partners and consultants who want to build AI capabilities into ERP deployments for their customers.

Software companies

Companies that want to create their own ERP-specific AI products using a reusable architecture.

Enterprises

Organizations that want greater control over their AI-agent infrastructure and source code.

The bigger idea

I don't see enterprise AI as simply putting a chatbot on top of an ERP.

The bigger opportunity is an AI intelligence layer for enterprise software.

That layer should be able to understand business context, interact with enterprise systems, use tools, follow permissions, execute workflows and provide traceability.

Noviz is my attempt to build that architecture.

ERPNext is the first implementation. The broader goal is an extensible ERP AI-agent platform that can be adapted to different enterprise systems.

I'm interested in hearing from developers working on AI agents, ERP integrations, RAG, tool calling, enterprise automation and agent architectures.

I'd especially like to know:

What would you add to an architecture like Noviz before deploying it in a real enterprise environment?