The J2 Innovations' blog

The home of smart buildings, smart equipment and IoT

FIN Intelligence: What Memoori Saw, and What It Can Do Now

Screenshot 2026-11-17 at 11.10.58 AM

FIN Intelligence is the generative AI layer built into FIN Framework. It lets engineers, integrators and operators work with a building in plain language, from asking about an alarm to building a Modbus connector.

On November 2, Memoori published a research note on J2 Innovations: FIN Intelligence Claims 80% Faster Tagging: J2's Gen AI. I've spent a lot of years in this industry, and it's always useful to see how an independent analyst reads what you're building. Memoori got a lot right.

The feature set in the note reflects what we described during our beta. Since then, FIN Intelligence has been fully released with FIN Framework 5.3, and it does considerably more. Below, I'll summarize what Memoori said and then walk through where FIN Intelligence stands today.

 

What Memoori said about J2 and FIN Framework

Memoori describes FIN Framework as a data integration platform rather than a full-stack smart building suite. It unifies diverse building systems through open protocols and Haystack tagging. OEMs build their own branded solutions on top of it, which saves controls and HVAC manufacturers from writing a modern building management interface from scratch.

That's a fair description, except out of the box, FIN Framework is a full suite of supervisory and control applications. That suite forms the foundation for the technology our partners leverage and build on.

 

FIN Intelligence as Memoori described it

Memoori notes that we launched FIN Intelligence at the Smart Buildings Show in London in October 2026, aimed at systems integrators and end users. The note describes three features built on large language models:

  • FIN Intelligent Tagging automates data tagging with the Project Haystack ontology. We've said it cuts project tagging time by 80%, which is where the note's headline comes from.
  • FIN Intelligent Optimization applies Plant Optimization to chiller plants, using real-time building data and predictions. We anticipate an energy efficiency gain of more than 10%.
  • FIN Intelligent Assistant helps detect faults such as lights left on, tracks energy trends, explains alarms and surfaces maintenance needs. In our first test, 78% of users saw immediate usability improvements.

Memoori also points out that FIN Intelligence runs on the edge and uses the cloud only when needed. Those three features are still part of the story, but FIN Intelligence has significantly continued to evolve (keep reading). 

 

What FIN Intelligence does now 

After a year of beta testing with customers worldwide, we officially launched FIN Intelligence globally in October 2026 with FIN Framework 5.3, called "The Intelligence Release." Customer feedback from the beta shaped what you see today.

The biggest change is scope. During beta, the assistant mostly answered questions. Now it understands where you are in FIN and can take action across commissioning, engineering, and operations.

Finding answers and getting around

  • Framework Q&A: Ask how to do something, such as registering an instance to Edge2Cloud, and it answers from the FIN documentation.
  • App navigation: It opens Schedules, DB Builder, Alarms and the other main apps for you.
  • Context awareness: It knows which app and the context of what you're looking at, so you can ask "Show me dashboard for City Center." 
  • Graphics: It opens project graphics, such as an AHU or floorplan view.
  • Project data: It answers questions about sites, floors, rooms, equipment, points, alarms and schedules.

Commissioning and engineering

  • Record management: It can be instructed to create, update, and delete records, which speeds up commissioning.
  • Auto-tagging: It proposes consistent Haystack tags based on point names, kinds, units and existing tags, and applies them once you accept.
  • Logic creation: It gathers your requirements in conversation, then opens Logic Builder to build the program.
  • Logic debugging: It catches problems the control engine won't flag, such as two programs writing to the same point or a routine missing a required tag.

Operations

  • Dashboards: Describe a layout or upload a sketch, and it builds the dashboard in FI Dashboards.
  • Alarm acknowledgement: It acknowledges specific alarms, with an optional message.
  • Point writes: It can be instructed to write values to writable points. For example, modify all the set points in a specific area by two degrees.
  • Schedules and calendars: It creates schedules bound to points and calendars, then opens the Schedules app.
  • Histories and charts: It reads and aggregates point histories and shows charts right in the chat.

We've also introduced FIN Intelligence Code for developers. It helps them build extensions, connectors and applications on top of FIN.

2026 Blogs-2-1

 

Memoori's view on competition and strategy

Memoori names Tridium's Niagara Framework as FIN's main competitor. Niagara has a larger installed base and a more mature partner ecosystem. Memoori says FIN has a more modern, semantically rich architecture that is arguably better positioned for data-centric buildings, but it needs to keep growing its footprint and partner network. I agree with both halves of that.

The note also compares FIN Intelligence to Delta's Building Canvas. Both target integrators and aim to reduce engineering effort. Memoori gives FIN the edge because it is an open, vendor-agnostic framework.

The most interesting part is Memoori's read on our channels. FIN Framework serves OEMs, while FIN Intelligence reaches systems integrators and building operators. Memoori sees four reasons this makes sense:

  • Integrators need analytics: SIs want to offer fault detection, analytics and optimization so they aren't competing on installation alone.
  • Recurring revenue: A service layer for the SI channel complements licensing from OEM partners.
  • A bigger market: An integrator can deploy FIN Intelligence on BACnet or Modbus systems that don't use OEM partner hardware.
  • A position next to Niagara: Many integrators already work in Niagara, and FIN Intelligence could run alongside those installations.

Memoori calls this dual-channel approach a smart way to cover the market without channel conflict. From my experience and the whole reason J2 Innovations was created, it comes down to something simpler. The people who commission and run buildings spend too much time on repetitive work, and FIN Intelligence is built to give that time back.

 

Where we go from here

Memoori's note is a good snapshot of where FIN Intelligence started. The full release goes much further. It moves from answering questions to doing real work across commissioning, engineering and day-to-day operations.

We'll keep building on what our beta customers taught us, and we'll keep FIN Framework open so it works with the systems already in your buildings. If you'd like to see what FIN Intelligence can do on your projects, visit the FIN Intelligence page or reach out to our team

B. Scott Muench

Scott joined J2 Innovations as a partner in 2011 and is now Vice President of Knowledge Excellence. He has a wide range of responsibilities, including evangelism, business development and training. Scott is well known as an industry expert in smart homes and smart buildings. He is a past president of ASHRAE, and is currently a board member for Project Haystack. Scott attended Clarkson University for Mechanical Engineering and graduated with a BS/Business in Organizational Innovation.

View all articles

Topics from this blog: Smart Buildings FIN Framework Industry Building Automation System BAS News AI FIN Intelligence

Back to all posts