There's a lot of noise around AI in ERP right now. Some of it is marketing. Some of it is genuinely useful. If you're running Odoo or evaluating it, you want to know which bits are actually ready to use in a real business today, which ones need careful setup, and what's coming that will change how you think about automation altogether.
This post covers the full picture: what Odoo AI does across every major module in v19, what the Australian context changes about the answer (GST, BAS, and compliance matter here), what Odoo's CEO Fabien Pinckaers announced about AI-assisted development in February 2026, and where Odoo v20 is taking things later this year. We've also been configuring AI inside Odoo since version 16, so we can tell you what the ROI actually looks like from client work, not from a vendor brochure.
How Odoo AI has evolved: from v17 to v20
Odoo's AI capabilities have changed substantially across recent versions. Before getting into the specifics, here's the version landscape, because where you are now shapes what's available to you:
| Version | Key AI additions | What it meant in practice |
|---|---|---|
| v17 | Basic content generation, early OCR | AI as a drafting aid. Useful for product descriptions and email templates. Not deeply embedded in business processes. |
| v18 | Improved document OCR, lead scoring in Sales, email drafting, expense product prediction | AI starts doing operational work. Invoice processing becomes materially faster. Lead scoring gives sales managers a prioritised pipeline view. |
| v19 | Dedicated AI app, AI Agents, AI automation rules, AI across Sales, Accounting, Documents, Knowledge, Livechat, Website | AI becomes a platform layer, not a feature list. Every module gets AI capability. Natural language database queries. Custom agents trained on your own data. |
| v20 (expected Sept 2026) | Agentic AI: autonomous, proactive workflow execution | Moves from responding to instructions to acting independently. The system checks, decides, and acts without being prompted. |
What AI features does Odoo 19 include?
Odoo 19 introduced a dedicated AI app that runs inside the modules you already use. It's not a bolt-on integration. The Odoo 19 release notes and the v19 Enterprise source code (which Auboros accesses as a certified Silver Partner) confirm AI integration across the following areas:
- Ask AI and natural language database queries: ask questions about your Odoo data in plain English, without navigating menus or building filters manually.
- AI Agents: configurable assistants you train on your own documents, knowledge base, and website content. They answer queries and, with the right setup, take actions inside Odoo.
- AI in Odoo Studio (automation rules): set up automations that use AI to evaluate conditions and update record fields without any custom code.
- AI Document Management: automatically sorts uploaded documents into the right folders based on content classification.
- AI in Accounting: AI-assisted drafting for payment follow-ups and collection letters, layered on top of the invoice OCR that has been in Odoo since v18.
- AI in Sales and CRM: automatic lead creation from emails, website interactions, and livechat, plus the pipeline probability scoring from v18.
- AI in Knowledge: AI text drafting and content improvement inside the Knowledge module.
- AI Website Builder: page creation and content drafting for businesses running their public website on Odoo.
- AI in Livechat and Helpdesk: suggested responses for agents in real time, plus ticket history summarisation.
AI across your Odoo modules: what it actually does
The feature list above is accurate but abstract. Here's what each area means for someone running a business in Australia.
Sales and CRM: smarter pipeline management
Lead scoring has been in Odoo since v18 and assigns probability percentages to deals based on historical close data. The model learns from your own pipeline, not a generic benchmark. A deal at 70% in one industry might behave completely differently from a deal at 70% in yours, and Odoo's scoring reflects your actual history.
In v19, automatic lead creation adds another layer. The system can identify and generate leads from website form submissions, livechat conversations, and incoming email threads without a sales rep manually creating a record. For teams managing high inquiry volume, this removes the admin gap between "someone contacted us" and "that contact is tracked in the pipeline."
Accounting: the clearest ROI for most businesses
Most businesses notice the accounting AI first, and for good reason. The document OCR engine reads uploaded supplier invoices, receipts, and bills, and pre-fills vendor name, ABN (where present), invoice total, GST amount, and due date. In an Australian context, getting the GST extraction right matters directly for your BAS (Business Activity Statement) accuracy. One wrongly coded line on a supplier invoice can create a reconciliation issue you won't find until BAS time.
If your accounts payable team is processing a high volume of supplier invoices, most of that data entry disappears once the system has learned your regular vendors. The time saving compounds quickly. Finance staff spending two hours a day on manual invoice entry get most of that time back.
V19 also adds AI-assisted drafting for overdue payment reminders and collection letters. For businesses that let aged receivables sit because nobody has time to write the follow-up, this removes the friction.
Document Management: intelligent sorting, without manual filing
Odoo's AI document sorting automatically classifies uploaded files and routes them to the right location based on content. Bills go to the Bills folder. Contracts go to Contracts. HR documents go to HR. For businesses that have a shared inbox or document upload process and rely on someone manually sorting what comes in, this is where you recover hidden administrative hours.
The AI document integration with accounting (a separate module in v19 Enterprise) goes a step further, linking sorted documents directly to accounting workflows so bills are ready for review rather than sitting in a queue waiting to be filed first.
AI Automation Rules: powerful without a developer
If your business doesn't have a dedicated development team, this is the feature worth paying close attention to. In Odoo's automation configuration, you can write rules in plain language that the system interprets and applies to your data. An automation rule might say: when a new enquiry comes in, use AI to assess the description and assign the opportunity to the right sales team based on what they're asking about. No code. No developer. Just a rule you configure in the settings.
The line between "using Odoo AI" and "building something custom" is starting to disappear. Workflows that used to require a developer to build can now be configured by a business analyst or operations manager with the right knowledge of the platform.
Ask AI and natural language queries: the quietest productivity gain
The ability to type "show me all overdue invoices from Queensland customers over $10,000" and get a filtered result immediately is practical and doesn't need configuration. Finance managers and operations staff who know what they want but often have to ask someone else to build the report will use this constantly. It lowers the cost of getting to your own data.
Knowledge and internal tools: keeping your process library current
For businesses using Odoo Knowledge as their operations manual or training resource, AI assists with drafting and improving articles. Most knowledge bases go stale because updating them takes effort nobody has time for. If the effort of updating a procedure becomes "open the article, click improve, review the suggestion," more people will actually do it.
Livechat and customer support: faster responses, less effort
For businesses running Odoo Livechat, AI suggests responses to common queries in real time. For Helpdesk, it summarises the history of an ongoing ticket when an agent picks it up. Both reduce time to a useful response for the customer, and reduce the cognitive load on the agent handling the interaction.
What real ROI looks like
We've been configuring AI features inside Odoo since version 16, across implementations in Queensland, New South Wales, and Victoria. The return varies by business, but the patterns are consistent. Accounts payable automation delivers the fastest visible result, typically within the first month of go-live. Businesses that were spending significant staff time on manual invoice entry see that time drop substantially, and the accuracy of their GST coding improves because the AI doesn't make transcription errors.
Beyond invoice processing, the more interesting ROI comes from automations that businesses couldn't afford to build before. Agentic workflows that trigger supplier communications based on inventory thresholds. AI agents trained on product catalogues that handle first-level customer queries without a human in the loop. Lead qualification rules that route high-intent enquiries to senior sales staff automatically. None of these are theoretical. They're things we've built in production for Queensland clients.
"We've been configuring AI inside Odoo since v16, well before it became a marketing headline. The ROI is real and the business cases are genuinely only limited by how well you understand your own processes. We've built some incredibly useful things for clients: invoice automation running overnight without anyone touching it, agentic workflows triggering supplier communications based on stock movements, AI agents handling live customer queries from a product catalogue. We're genuinely excited about what v20 will bring. As AI matures in the Odoo ecosystem, the move towards truly autonomous workflows is going to open up use cases we haven't built yet."
Bill Alvarez, Practice Manager, Auboros
According to MYOB's 2025 ERP Trends Report, 45% of Australian decision-makers say disconnected systems limit their business growth. The businesses getting the most from Odoo AI are the ones where everything, finance, inventory, sales, support, sits inside a single system. The AI has richer data to work with, the automations don't have to cross system boundaries, and the natural language queries can see the whole picture.
What this means for Australian businesses specifically
Odoo's Australian localisation covers GST, BAS reporting, and Single Touch Payroll (STP) Phase 2. The AI features don't replace any of that, but they make it easier to feed those reports accurately. Where this actually shows up:
- Invoice processing with GST: when Odoo AI reads a supplier invoice, it extracts the GST amount and codes it to the right tax account. If your vendor and tax mapping are correct, the pre-filled entries include the right GST treatment. This matters for BAS accuracy.
- Natural language queries for BAS prep: instead of building a custom report, you can ask "show me all GST collected this quarter" directly. It doesn't generate the BAS itself, but reviewing the underlying data before lodgement becomes much faster.
- AI Agents for compliance queries: train an agent on your ATO correspondence, payroll policies, and internal procedures, and it can handle routine team questions about process without someone senior stepping in each time.
- STP and payroll compliance: Odoo's Australian payroll module handles Single Touch Payroll Phase 2 reporting. AI doesn't touch the STP submission itself, but it assists with the surrounding HR and payroll workflow, drafting communications and summarising employee records.
For current ATO compliance requirements, ato.gov.au is the authoritative source. Odoo's own Australian localisation documentation covers how these are implemented in the platform.
Claude Code and Odoo.sh: what Fabien Pinckaers announced in February 2026
This is the announcement that generated the most discussion in the Odoo developer community in early 2026. In February, Odoo CEO Fabien Pinckaers announced that vibe coding is being officially integrated into Odoo.sh with Claude Code as the AI development tool of choice.
What this means practically: developers using Odoo.sh can use Claude Code directly within the platform to generate custom Odoo modules by describing what they want in natural language. Claude Code reads the entire project context, including existing models, manifest files, and inheritance chains, and generates multi-file changes simultaneously. It's not a generic code generator. It understands Odoo's framework patterns because the model has been trained extensively on Odoo's open-source codebase, which is a genuine advantage over trying to use a general-purpose AI assistant for Odoo development.
Pinckaers describes the approach as three elements: standard Odoo applications as the foundation, rapid AI-assisted customisation as the mechanism, and Odoo.sh as the platform that keeps deployment safe. The claim is that custom features that previously took weeks of development can be produced in minutes.
Community reaction has been engaged and honestly divided. Some developers see this as a step-change in how quickly small teams can customise Odoo for clients. Others raise legitimate concerns about unreviewed AI-generated code in live business environments, drawing comparisons to early Odoo Studio experiences with code quality and upgrade path stability. The position we'd take: the AI generates faster, but someone who understands Odoo still needs to review what comes out before it goes near production data.
For Odoo clients in Australia, the practical implication is that customisation is becoming more accessible. Features that previously required a larger development engagement can potentially be delivered faster and at lower cost. That's good news for businesses who've wanted specific workflows but couldn't justify the investment.
What about Odoo v18? And what we've built for clients on it
If you're on Odoo 18, which many Australian businesses are given it reached stable release in late 2024, you're not waiting on the sidelines. Version 18 includes document OCR for invoice processing, text generation across the platform, email drafting, lead scoring in CRM, and an expense product prediction feature. The dedicated AI app with Agents is a v19 addition, but the most commonly used day-to-day AI features are available right now on v18.
We've also built native AI capabilities directly into v18 implementations for several Queensland clients, going beyond the standard feature set. Document classification rules, automated purchase order matching, and workflow automations triggered by AI-detected conditions are all things we've configured within v18 without waiting for a version upgrade. If you're on v18 and want to understand what's accessible now versus what requires moving to v19, it's worth a direct conversation rather than going by the published feature list alone.
Auboros is certified in Odoo v18 and works with clients upgrading to v19. If you're planning an implementation or upgrade, timing your version choice to align with which AI capabilities you want access to is worth factoring in early.
What's coming in Odoo v20: agentic AI
Odoo v20 is expected at Odoo Experience in September 2026 in Brussels, and the anticipated headline feature is agentic AI. The distinction from v19 is significant: v19 AI responds when asked. V20 AI is expected to act without being asked, proactively executing workflows based on conditions it monitors itself.
The practical difference is the gap between a smart calculator and an employee who checks inventory every morning, files the reorder automatically, and emails the supplier before you knew there was a short. For manufacturing businesses, agentic AI embedded in production scheduling could adjust work orders in real time based on machine downtime data, supplier delays, and demand signals simultaneously. For distribution businesses, it could monitor stock movements and trigger purchase orders or customer notifications autonomously.
These are roadmap directions, not confirmed feature announcements. Odoo makes official product announcements at Odoo Experience each year. What's clear is that the v19 AI platform layer is being built specifically as the foundation for this kind of autonomous operation. Getting your Odoo implementation AI-ready now, with clean data, well-structured automations, and the right module configuration, is what makes upgrading to v20 a natural progression rather than a disruptive project.
Do you need Enterprise to use Odoo AI features?
Yes. The AI app and most AI features require an Enterprise subscription. Some features also require configuring an AI provider API key, typically OpenAI, or using Odoo's IAP (In-App Purchase) credits available through your Odoo.com subscription. Your implementation partner handles this as part of the setup.
One thing worth being clear on: Odoo Enterprise source code is accessible to all Enterprise subscribers, not just partners. So the question of whether to bring in a partner for AI configuration or handle it internally is a capability question, not an access question. If your internal team has the technical depth, a lot of the AI setup is something you can manage yourselves. If it's not a core competency, a partner will get you there faster and with fewer configuration mistakes to unpick later.
There are community-built approaches for adding AI to Odoo Community Edition using workflow tools and open-source integrations, but these carry more maintenance overhead. For most Australian mid-market businesses, the Enterprise route is the practical path. The Auboros implementation packages page covers what's included at each tier.
Frequently asked questions
Is Odoo AI available for Australian businesses?
Yes. All Odoo AI features are available globally with no geographic restrictions. Australian Enterprise subscribers access them the same way as anyone else. Setup requires configuring an AI provider API key or IAP credits, which your implementation partner handles as part of onboarding.
Does Odoo AI help with BAS and GST processing?
Indirectly, yes. Odoo AI improves invoice data entry accuracy, which feeds directly into your BAS (Business Activity Statement) figures. Natural language queries let you review GST-coded transactions before lodgement without building a custom report. Odoo's Australian localisation handles the BAS report structure itself. The AI assists the inputs, not the report. For current BAS and GST requirements, ato.gov.au is the authoritative source.
Can Odoo AI read and process supplier invoices automatically?
Yes. Odoo's invoice OCR reads uploaded bills and pre-fills vendor details, amounts, GST lines, and due dates. Accuracy improves over time as the system learns your regular vendors. You still review and approve before posting. Available in both v18 and v19. V19 adds automatic document sorting on top of processing.
What does Claude Code in Odoo.sh mean for my business?
It means customisations that previously required significant development time are becoming faster and less expensive to deliver. Fabien Pinckaers announced in February 2026 that Claude Code is being integrated into Odoo.sh for AI-assisted module development. Features you've wanted but couldn't justify the build cost for are increasingly within reach. The code still needs review by someone who knows Odoo before it goes to production, but the build time is shrinking.
When is the right time to add AI features to an Odoo implementation?
Start with a stable core first. AI features work best when your data is clean, your chart of accounts is correct, and your team understands the basics. Most Auboros clients look at AI configuration in phase two, once the core go-live is settled and the team knows what they actually want automated. Layering AI on top of a setup that hasn't been properly configured creates more problems than it solves.
Is Odoo AI ready to replace an accountant or finance manager?
No, and it's not designed to. What it handles is the volume work: data entry, routine lookups, repetitive drafting. Judgment calls, exception handling, and the professional accountability that comes with BAS lodgement and STP payroll still need a person. The value is in returning hours of manual work to your finance team so they can focus on the work that actually needs thinking.
Thinking about AI as part of your Odoo implementation?
We've been building Odoo AI configurations for Queensland and New South Wales businesses since v16. We know what delivers real results and what needs more setup than most businesses are prepared for. Getting the architecture right from the start is what makes AI genuinely useful rather than a feature nobody ends up using.
If you're planning an Odoo implementation or upgrade and want a straight conversation about where AI fits in your specific situation, book a free consultation. No obligation, just an honest assessment of what makes sense for your business and your timeline.
There's a lot of noise around AI in ERP right now. Some of it is marketing. Some of it is genuinely useful. If you're running Odoo or evaluating it, you want to know which bits are actually ready to use in a real business today, which ones need careful setup, and what's coming that will change how you think about automation altogether.
This post covers the full picture: what Odoo AI does across every major module in v19, what the Australian context changes about the answer (GST, BAS, and compliance matter here), what Odoo's CEO Fabien Pinckaers announced about AI-assisted development in February 2026, and where Odoo v20 is taking things later this year. We've also been configuring AI inside Odoo since version 16, so we can tell you what the ROI actually looks like from client work, not from a vendor brochure.
How Odoo AI has evolved: from v17 to v20
Odoo's AI capabilities have changed substantially across recent versions. Before getting into the specifics, here's the version landscape, because where you are now shapes what's available to you:
| Version | Key AI additions | What it meant in practice |
|---|---|---|
| v17 | Basic content generation, early OCR | AI as a drafting aid. Useful for product descriptions and email templates. Not deeply embedded in business processes. |
| v18 | Improved document OCR, lead scoring in Sales, email drafting, expense product prediction | AI starts doing operational work. Invoice processing becomes materially faster. Lead scoring gives sales managers a prioritised pipeline view. |
| v19 | Dedicated AI app, AI Agents, AI automation rules, AI across Sales, Accounting, Documents, Knowledge, Livechat, Website | AI becomes a platform layer, not a feature list. Every module gets AI capability. Natural language database queries. Custom agents trained on your own data. |
| v20 (expected Sept 2026) | Agentic AI: autonomous, proactive workflow execution | Moves from responding to instructions to acting independently. The system checks, decides, and acts without being prompted. |
What AI features does Odoo 19 include?
Odoo 19 introduced a dedicated AI app that runs inside the modules you already use. It's not a bolt-on integration. The Odoo 19 release notes and the v19 Enterprise source code (which Auboros accesses as a certified Silver Partner) confirm AI integration across the following areas:
- Ask AI and natural language database queries: ask questions about your Odoo data in plain English, without navigating menus or building filters manually.
- AI Agents: configurable assistants you train on your own documents, knowledge base, and website content. They answer queries and, with the right setup, take actions inside Odoo.
- AI in Odoo Studio (automation rules): set up automations that use AI to evaluate conditions and update record fields without any custom code.
- AI Document Management: automatically sorts uploaded documents into the right folders based on content classification.
- AI in Accounting: AI-assisted drafting for payment follow-ups and collection letters, layered on top of the invoice OCR that has been in Odoo since v18.
- AI in Sales and CRM: automatic lead creation from emails, website interactions, and livechat, plus the pipeline probability scoring from v18.
- AI in Knowledge: AI text drafting and content improvement inside the Knowledge module.
- AI Website Builder: page creation and content drafting for businesses running their public website on Odoo.
- AI in Livechat and Helpdesk: suggested responses for agents in real time, plus ticket history summarisation.
AI across your Odoo modules: what it actually does
The feature list above is accurate but abstract. Here's what each area means for someone running a business in Australia.
Sales and CRM: smarter pipeline management
Lead scoring has been in Odoo since v18 and assigns probability percentages to deals based on historical close data. The model learns from your own pipeline, not a generic benchmark. A deal at 70% in one industry might behave completely differently from a deal at 70% in yours, and Odoo's scoring reflects your actual history.
In v19, automatic lead creation adds another layer. The system can identify and generate leads from website form submissions, livechat conversations, and incoming email threads without a sales rep manually creating a record. For teams managing high inquiry volume, this removes the admin gap between "someone contacted us" and "that contact is tracked in the pipeline."
Accounting: the clearest ROI for most businesses
Most businesses notice the accounting AI first, and for good reason. The document OCR engine reads uploaded supplier invoices, receipts, and bills, and pre-fills vendor name, ABN (where present), invoice total, GST amount, and due date. In an Australian context, getting the GST extraction right matters directly for your BAS (Business Activity Statement) accuracy. One wrongly coded line on a supplier invoice can create a reconciliation issue you won't find until BAS time.
If your accounts payable team is processing a high volume of supplier invoices, most of that data entry disappears once the system has learned your regular vendors. The time saving compounds quickly. Finance staff spending two hours a day on manual invoice entry get most of that time back.
V19 also adds AI-assisted drafting for overdue payment reminders and collection letters. For businesses that let aged receivables sit because nobody has time to write the follow-up, this removes the friction.
Document Management: intelligent sorting, without manual filing
Odoo's AI document sorting automatically classifies uploaded files and routes them to the right location based on content. Bills go to the Bills folder. Contracts go to Contracts. HR documents go to HR. For businesses that have a shared inbox or document upload process and rely on someone manually sorting what comes in, this is where you recover hidden administrative hours.
The AI document integration with accounting (a separate module in v19 Enterprise) goes a step further, linking sorted documents directly to accounting workflows so bills are ready for review rather than sitting in a queue waiting to be filed first.
AI Automation Rules: powerful without a developer
If your business doesn't have a dedicated development team, this is the feature worth paying close attention to. In Odoo's automation configuration, you can write rules in plain language that the system interprets and applies to your data. An automation rule might say: when a new enquiry comes in, use AI to assess the description and assign the opportunity to the right sales team based on what they're asking about. No code. No developer. Just a rule you configure in the settings.
The line between "using Odoo AI" and "building something custom" is starting to disappear. Workflows that used to require a developer to build can now be configured by a business analyst or operations manager with the right knowledge of the platform.
Ask AI and natural language queries: the quietest productivity gain
The ability to type "show me all overdue invoices from Queensland customers over $10,000" and get a filtered result immediately is practical and doesn't need configuration. Finance managers and operations staff who know what they want but often have to ask someone else to build the report will use this constantly. It lowers the cost of getting to your own data.
Knowledge and internal tools: keeping your process library current
For businesses using Odoo Knowledge as their operations manual or training resource, AI assists with drafting and improving articles. Most knowledge bases go stale because updating them takes effort nobody has time for. If the effort of updating a procedure becomes "open the article, click improve, review the suggestion," more people will actually do it.
Livechat and customer support: faster responses, less effort
For businesses running Odoo Livechat, AI suggests responses to common queries in real time. For Helpdesk, it summarises the history of an ongoing ticket when an agent picks it up. Both reduce time to a useful response for the customer, and reduce the cognitive load on the agent handling the interaction.
What real ROI looks like
We've been configuring AI features inside Odoo since version 16, across implementations in Queensland, New South Wales, and Victoria. The return varies by business, but the patterns are consistent. Accounts payable automation delivers the fastest visible result, typically within the first month of go-live. Businesses that were spending significant staff time on manual invoice entry see that time drop substantially, and the accuracy of their GST coding improves because the AI doesn't make transcription errors.
Beyond invoice processing, the more interesting ROI comes from automations that businesses couldn't afford to build before. Agentic workflows that trigger supplier communications based on inventory thresholds. AI agents trained on product catalogues that handle first-level customer queries without a human in the loop. Lead qualification rules that route high-intent enquiries to senior sales staff automatically. None of these are theoretical. They're things we've built in production for Queensland clients.
"We've been configuring AI inside Odoo since v16, well before it became a marketing headline. The ROI is real and the business cases are genuinely only limited by how well you understand your own processes. We've built some incredibly useful things for clients: invoice automation running overnight without anyone touching it, agentic workflows triggering supplier communications based on stock movements, AI agents handling live customer queries from a product catalogue. We're genuinely excited about what v20 will bring. As AI matures in the Odoo ecosystem, the move towards truly autonomous workflows is going to open up use cases we haven't built yet."
Bill Alvarez, Practice Manager, Auboros
According to MYOB's 2025 ERP Trends Report, 45% of Australian decision-makers say disconnected systems limit their business growth. The businesses getting the most from Odoo AI are the ones where everything, finance, inventory, sales, support, sits inside a single system. The AI has richer data to work with, the automations don't have to cross system boundaries, and the natural language queries can see the whole picture.
What this means for Australian businesses specifically
Odoo's Australian localisation covers GST, BAS reporting, and Single Touch Payroll (STP) Phase 2. The AI features don't replace any of that, but they make it easier to feed those reports accurately. Where this actually shows up:
- Invoice processing with GST: when Odoo AI reads a supplier invoice, it extracts the GST amount and codes it to the right tax account. If your vendor and tax mapping are correct, the pre-filled entries include the right GST treatment. This matters for BAS accuracy.
- Natural language queries for BAS prep: instead of building a custom report, you can ask "show me all GST collected this quarter" directly. It doesn't generate the BAS itself, but reviewing the underlying data before lodgement becomes much faster.
- AI Agents for compliance queries: train an agent on your ATO correspondence, payroll policies, and internal procedures, and it can handle routine team questions about process without someone senior stepping in each time.
- STP and payroll compliance: Odoo's Australian payroll module handles Single Touch Payroll Phase 2 reporting. AI doesn't touch the STP submission itself, but it assists with the surrounding HR and payroll workflow, drafting communications and summarising employee records.
For current ATO compliance requirements, ato.gov.au is the authoritative source. Odoo's own Australian localisation documentation covers how these are implemented in the platform.
Claude Code and Odoo.sh: what Fabien Pinckaers announced in February 2026
This is the announcement that generated the most discussion in the Odoo developer community in early 2026. In February, Odoo CEO Fabien Pinckaers announced that vibe coding is being officially integrated into Odoo.sh with Claude Code as the AI development tool of choice.
What this means practically: developers using Odoo.sh can use Claude Code directly within the platform to generate custom Odoo modules by describing what they want in natural language. Claude Code reads the entire project context, including existing models, manifest files, and inheritance chains, and generates multi-file changes simultaneously. It's not a generic code generator. It understands Odoo's framework patterns because the model has been trained extensively on Odoo's open-source codebase, which is a genuine advantage over trying to use a general-purpose AI assistant for Odoo development.
Pinckaers describes the approach as three elements: standard Odoo applications as the foundation, rapid AI-assisted customisation as the mechanism, and Odoo.sh as the platform that keeps deployment safe. The claim is that custom features that previously took weeks of development can be produced in minutes.
Community reaction has been engaged and honestly divided. Some developers see this as a step-change in how quickly small teams can customise Odoo for clients. Others raise legitimate concerns about unreviewed AI-generated code in live business environments, drawing comparisons to early Odoo Studio experiences with code quality and upgrade path stability. The position we'd take: the AI generates faster, but someone who understands Odoo still needs to review what comes out before it goes near production data.
For Odoo clients in Australia, the practical implication is that customisation is becoming more accessible. Features that previously required a larger development engagement can potentially be delivered faster and at lower cost. That's good news for businesses who've wanted specific workflows but couldn't justify the investment.
What about Odoo v18? And what we've built for clients on it
If you're on Odoo 18, which many Australian businesses are given it reached stable release in late 2024, you're not waiting on the sidelines. Version 18 includes document OCR for invoice processing, text generation across the platform, email drafting, lead scoring in CRM, and an expense product prediction feature. The dedicated AI app with Agents is a v19 addition, but the most commonly used day-to-day AI features are available right now on v18.
We've also built native AI capabilities directly into v18 implementations for several Queensland clients, going beyond the standard feature set. Document classification rules, automated purchase order matching, and workflow automations triggered by AI-detected conditions are all things we've configured within v18 without waiting for a version upgrade. If you're on v18 and want to understand what's accessible now versus what requires moving to v19, it's worth a direct conversation rather than going by the published feature list alone.
Auboros is certified in Odoo v18 and works with clients upgrading to v19. If you're planning an implementation or upgrade, timing your version choice to align with which AI capabilities you want access to is worth factoring in early.
What's coming in Odoo v20: agentic AI
Odoo v20 is expected at Odoo Experience in September 2026 in Brussels, and the anticipated headline feature is agentic AI. The distinction from v19 is significant: v19 AI responds when asked. V20 AI is expected to act without being asked, proactively executing workflows based on conditions it monitors itself.
The practical difference is the gap between a smart calculator and an employee who checks inventory every morning, files the reorder automatically, and emails the supplier before you knew there was a short. For manufacturing businesses, agentic AI embedded in production scheduling could adjust work orders in real time based on machine downtime data, supplier delays, and demand signals simultaneously. For distribution businesses, it could monitor stock movements and trigger purchase orders or customer notifications autonomously.
These are roadmap directions, not confirmed feature announcements. Odoo makes official product announcements at Odoo Experience each year. What's clear is that the v19 AI platform layer is being built specifically as the foundation for this kind of autonomous operation. Getting your Odoo implementation AI-ready now, with clean data, well-structured automations, and the right module configuration, is what makes upgrading to v20 a natural progression rather than a disruptive project.
Do you need Enterprise to use Odoo AI features?
Yes. The AI app and most AI features require an Enterprise subscription. Some features also require configuring an AI provider API key, typically OpenAI, or using Odoo's IAP (In-App Purchase) credits available through your Odoo.com subscription. Your implementation partner handles this as part of the setup.
One thing worth being clear on: Odoo Enterprise source code is accessible to all Enterprise subscribers, not just partners. So the question of whether to bring in a partner for AI configuration or handle it internally is a capability question, not an access question. If your internal team has the technical depth, a lot of the AI setup is something you can manage yourselves. If it's not a core competency, a partner will get you there faster and with fewer configuration mistakes to unpick later.
There are community-built approaches for adding AI to Odoo Community Edition using workflow tools and open-source integrations, but these carry more maintenance overhead. For most Australian mid-market businesses, the Enterprise route is the practical path. The Auboros implementation packages page covers what's included at each tier.
Frequently asked questions
Is Odoo AI available for Australian businesses?
Yes. All Odoo AI features are available globally with no geographic restrictions. Australian Enterprise subscribers access them the same way as anyone else. Setup requires configuring an AI provider API key or IAP credits, which your implementation partner handles as part of onboarding.
Does Odoo AI help with BAS and GST processing?
Indirectly, yes. Odoo AI improves invoice data entry accuracy, which feeds directly into your BAS (Business Activity Statement) figures. Natural language queries let you review GST-coded transactions before lodgement without building a custom report. Odoo's Australian localisation handles the BAS report structure itself. The AI assists the inputs, not the report. For current BAS and GST requirements, ato.gov.au is the authoritative source.
Can Odoo AI read and process supplier invoices automatically?
Yes. Odoo's invoice OCR reads uploaded bills and pre-fills vendor details, amounts, GST lines, and due dates. Accuracy improves over time as the system learns your regular vendors. You still review and approve before posting. Available in both v18 and v19. V19 adds automatic document sorting on top of processing.
What does Claude Code in Odoo.sh mean for my business?
It means customisations that previously required significant development time are becoming faster and less expensive to deliver. Fabien Pinckaers announced in February 2026 that Claude Code is being integrated into Odoo.sh for AI-assisted module development. Features you've wanted but couldn't justify the build cost for are increasingly within reach. The code still needs review by someone who knows Odoo before it goes to production, but the build time is shrinking.
When is the right time to add AI features to an Odoo implementation?
Start with a stable core first. AI features work best when your data is clean, your chart of accounts is correct, and your team understands the basics. Most Auboros clients look at AI configuration in phase two, once the core go-live is settled and the team knows what they actually want automated. Layering AI on top of a setup that hasn't been properly configured creates more problems than it solves.
Is Odoo AI ready to replace an accountant or finance manager?
No, and it's not designed to. What it handles is the volume work: data entry, routine lookups, repetitive drafting. Judgment calls, exception handling, and the professional accountability that comes with BAS lodgement and STP payroll still need a person. The value is in returning hours of manual work to your finance team so they can focus on the work that actually needs thinking.
Thinking about AI as part of your Odoo implementation?
We've been building Odoo AI configurations for Queensland and New South Wales businesses since v16. We know what delivers real results and what needs more setup than most businesses are prepared for. Getting the architecture right from the start is what makes AI genuinely useful rather than a feature nobody ends up using.
If you're planning an Odoo implementation or upgrade and want a straight conversation about where AI fits in your specific situation, book a free consultation. No obligation, just an honest assessment of what makes sense for your business and your timeline.
Start writing here...