Another Acumatica Summit done, and this year felt different. New venue, new leadership, and AI that has finally moved from promise to product. Here's my honest take on what stood out, what deserves scepticism, and what you should genuinely be paying attention to as an Acumatica customer or partner.
Seattle: A Better Home for This Conference
The move from the Wynn in Las Vegas to the Seattle Convention Center was, for me, a clear upgrade. Not because Vegas isn't fun — it is — but because walking through a casino that smells of cigarettes at 8am to get to an ERP conference was never a great look. Seattle felt more professional, more business-focused, and the venue itself genuinely suited the event.
The multi-floor layout encouraged the kind of spontaneous networking that the Wynn's sprawling single-level setup made difficult. You'd pass people on escalators, stop for a five-minute conversation between sessions, and move on. That organic flow of conversation — between partners, customers, and the Acumatica team — is really where a lot of the value of these events lives. Acumatica has already confirmed the Seattle Convention Center for next year, and I think that's the right call.
The Big Theme: Your ERP as a Digital Replica of Your Business
The overarching message from Acumatica this year was a shift in how to think about what an ERP actually is. Rather than a data store you interrogate after the fact, the vision being presented is of an ERP as a live digital replica of your business — capturing not just what happened, but driving what happens next.
It's a phrase that lands well because it's accurate. When an ERP is implemented and maintained properly, it should reflect every meaningful process and transaction in the business in real time. The numbers Acumatica shared to back this up are hard to argue with: over $500 billion transacted through the platform, 60 million product shipments, 280,000 daily unique users, and over 10 billion API calls per year. That last number is particularly telling — it signals just how connected this ecosystem is becoming, not just as a standalone system but as the hub that everything else plugs into.
AI: Finally Shipping, and Mostly Practical
AI was, unsurprisingly, the dominant theme. But what was different this year is that we moved from roadmap promises to actual features being shipped in 2026 R1. And more importantly, most of what was demonstrated had a clear, practical use case rather than just being technically impressive.
A few highlights that stood out:
Opportunity sentiment analysis. The ability to pull together all the activity on a sales opportunity — emails, notes, interactions — and get an AI-generated summary of where it stands, what the next steps are, and what the customer sentiment looks like. Useful on its own, but the more interesting point is that the underlying model is extensible through AI Studio, so the same logic can be applied across other modules.
AI Studio. This is the orchestration layer that makes the rest of it work. Low-code or no-code, it lets you configure your own AI assistants within your Acumatica tenant, plug in whatever LLM you want to use — OpenAI, Claude, Gemini, Azure — and set your own system prompts. The ceiling on what's possible here is genuinely high, and the hackathon at Summit gave a good demonstration of that: the winning project let someone on the road verbally dictate notes, which were then transcribed, turned into action items, and used to update customer and project records automatically.
Customer portal AI assistant. Customers being able to ask about their orders, invoices, and backorders through a self-service AI interface is a meaningful shift for customer service teams that currently field a high volume of routine queries.
MCP architecture. This is the one I haven't seen discussed much elsewhere, and it deserves attention. Model Context Protocol is essentially a standard that allows AI to connect to external tools and data sources — a universal adapter, if you like. Acumatica demonstrated this through a field services use case where the AI assistant could access current inventory, scheduling, customer history, and live weather data simultaneously. The result was a scheduling recommendation that flagged not just a calendar clash but an actual delivery risk due to weather conditions at a specific location. That kind of contextual reasoning — pulling from inside and outside the ERP in one response — is a meaningful step beyond what AI could do when it was confined to the system itself.
One note of caution on all of this: AI is only as good as the data you feed it. If you're migrating from a legacy system with messy, incomplete, or inconsistent data, AI will give you messy, inconsistent answers — confidently. Data quality isn't a technical concern that sits below AI strategy. It is the AI strategy.
There's also a practical concern around Acumatica's transaction-based licensing model. If you open up an AI assistant on the customer portal, you're now at the mercy of your customers' query behaviour for your transaction count. That could get expensive quickly, and it's something worth modelling out before you turn it on.
Insight XL: Interesting, But I'll Wait
Acumatica announced Insight XL, their native live Excel connection — essentially a direct competitor to Velixo. Click the Excel icon on any report, choose the live link option, and you get a spreadsheet where every field is a live connection back to Acumatica. There are built-in functions for things like chart of accounts lookups and ending balances.
My honest reaction: good direction, but not there yet. Velixo is significantly more capable, particularly around writeback functionality — the ability to push data back into Acumatica from Excel, which is invaluable for things like budget and forecast entry from finance teams or field sales. Insight XL doesn't appear to have that.
If it's included in the standard licence, it's worth exploring as a lightweight option. If it comes with an additional charge — and given that it's been given a distinct product name, I'd expect it will — I'll stick with Power Query and a GI connector. It's not meaningfully different in outcome, and it's already working.
Product Updates Worth Noting
A few other items that caught my attention:
Order orchestration assistant. Rules-based automated fulfillment routing that takes the decision-making burden off customer service reps and optimises across warehouses based on current inventory. Practical, time-saving, and a natural candidate for AI enhancement down the line — imagine factoring in real-time weather or carrier disruptions through MCP.
Shop floor kiosk. A dedicated UI for the shop floor that makes it genuinely easy for operatives to log progress and scan transactions. This matters because the shop floor is typically where ERP adoption dies. If reporting requires navigating a full ERP screen, it won't happen consistently. A purpose-built interface removes that barrier, and the data that flows from it — scrap, actual vs. planned progress, variances — underpins everything from accurate costing to production scheduling.
Warehouse management improvements. The ability to change pick locations mid-pick and handle partial picks more cleanly are useful incremental improvements. That said, my honest view is that Acumatica should either acquire a strong ISV solution in this space or focus warehouse management development effort on closing the gaps that matter most — projects integration, for instance, is still missing. Iterating on things that third-party ISVs have already solved years ago is not the best use of development resource.
Vendor portal. Coming in the roadmap, and genuinely exciting. Vendors being able to log in and update their own expected receipt dates and shipping status removes a significant administrative burden from supply chain teams. This is the kind of feature that has a direct, measurable impact on planning accuracy.
Teams integration via business events. Practical. A lot of businesses now operate significantly through Teams, and being able to trigger notifications there from business events in Acumatica closes a gap that a lot of customers have been asking about.
Leadership Transition
Ali Jani stepped back from his Chief Product Officer role and handed over to Jon Pollock. The handover was handled well — Ali brought the energy he always brings on stage, introduced Pollock with genuine warmth, and Pollock came across as already well across the product and direction. Transitions like this are always a moment of uncertainty for a platform's customers and partners. From what was presented, it looks like continuity is the intention rather than a change of direction.
What's Coming in 2026
The roadmap slides included warehouse bin replenishment, advanced kitting improvements, further development of the Advanced Planning and Scheduling module, product configurator improvements, AI for e-commerce and planning, customer portal modernisation, and the vendor portal mentioned above. Manufacturing customers in particular should be watching the APS and shop floor developments closely — when those capabilities mature and connect properly to accurate shop floor data, the planning and costing picture becomes significantly more useful.
The Bottom Line
Acumatica Summit 2026 felt like a platform at a confident moment. The AI features being shipped are practical rather than performative, the MCP integration points to a genuinely interesting direction for how ERP connects to the broader technology stack, and the Seattle venue gave the whole event a more professional, focused feel.
The things to watch are the same things that were always true: data quality underpins everything, AI doesn't fix bad processes, and the value of any of these features depends entirely on whether people are actually using the system properly. The technology is getting better. The fundamentals haven't changed.