Fenix, the multifamily AI personal assistant — Inside the Funnel with CJ Donio

In this episode of Inside the Funnel, host Hannah O'Leary sits down with CJ, Director of Product for AI Workflows and Insights, to talk about the personal AI assistant he's building for Funnel teams. CJ shares what carries over from building generative AI across edtech and other industries, why keeping a human in the loop matters, and how the assistant can query your data, help you act on it, and eventually tell you where to focus. Along the way, they unpack the buzzwords operators keep hearing, like MCP and orchestration, and what they actually mean for running your business.

You'll hear from:

Hannah O'Leary, Communications director, Funnel

CJ Donio, Director of Product for AI, Workflows and Insights, Funnel

In this episode:

• Lessons that carry over - building trustworthy AI on customers' data, keeping the UX simple, and always asking if a feature genuinely helps renters and leasing agents.
• A personal AI assistant for Funnel teams - query your data ("how many people didn't show up for a tour today?"), then act on it by assigning tasks or sending communications.
• Human in the loop - why the goal is to give teams a superpower, not remove the human, so agents can focus on relationships and one-to-one follow-up.
• MCP, explained - what the Model Context Protocol is ("an API for APIs"), why it's lightweight, and how it lets AI pull context from your PMS, partners, and Funnel.
• Granular permissions - how MCP connectors respect your existing Funnel login and access, down to the individual community level.
• Conversational BI for execs and data teams - querying Funnel data inside tools like Claude, ChatGPT, or Gemini and turning it into tables and charts.
• Bring your own app stack - building a robust integration platform so operators can keep the partners they love.
• Ramping new team members faster - using SOPs, KB content, Insights call coaching, custom scorecards, and adjustable automation to support newer teams.
• A look at the crystal ball - generative UI, where you describe the job to be done and the interface builds itself around it.
• Getting to know CJ - music, drums, a two-year-old, and why the people are his favorite part of Funnel.

Follow:

CJ Donio on LinkedIn

Hannah O'Leary on LinkedIn

Subscribe:

• Multifamily Unpacked podcast
• Multifamily Unpacked newsletter

Episode transcript:

0:00 Fenix, the multifamily AI personal assistant - Inside the Funnel with CJ Donio

CJ Donio: For the last four or five years everyone's trying to solve, which is how do I get my own personal assistant that knows me, knows my workflows, knows the different tools I use, and it could be my assistant I delegate work to.

The way we're talking about this, the way we're designing it is very much like we give you a superpower, but there's always that human in the loop.

Some of the stuff we're talking about that's really exciting is you basically can query data. You could say, "Hey, how many people didn't show up for a tour today?" And it's going to respond with that data, and then you can actually act on that data. This allows you to almost be the coach and tell your players where to be on the field.

I'm super excited to see how that evolves to the point where you basically can delegate a bunch of things that would stress you out, and now you can focus on building relationships and following up with renters one-on-one where your time is most valuable.

0:57 Introduction

Hannah O'Leary: Welcome back to Inside the Funnel, where we sit down with the people behind the product. I'm Hannah O'Leary producer of the show, and today I'm joined by CJ Donio, Director of Product for AI, Workflows, and Insights here at Funnel.

CJ has spent years building generative AI across industries. Today we dig into Fenix, which is part personal assistant, part always-on support team, part business analyst.

Fenix doesn't just answer; it acts, handling product and SOP questions, scheduling tours, setting reminders, and surfacing answers from your own portfolio data. You can use it inside Funnel or in the LLM of your choice.

We'll also unpack current AI buzzwords like MCP and orchestration. There's a lot to cover. Let's get into it.

01:58 Meet CJ Donio, Funnel Director of Product for AI, Workflows and Insights

Hannah O'Leary: CJ, thank you for being here today.

Before Funnel, you spent years building generative AI and platform products, from brand voice and APIs at Jasper to authentication and partner platforms at Divi, plus edtech. Now you're the director of product and AI workflows and insights in multifamily here at Funnel.

Obviously, when you start in a new industry, there are things that you bring from your past jobs into what you're doing here. What lessons from those worlds, especially around building trustworthy generative AI on top of customers' data, carry over into building AI for an industry that's ultimately about helping renters find a place that they call home?

And then what's genuinely different? What stands out to you as unique and sets multifamily apart from those other industries that you've worked in?

CJ Donio: Yeah, that's a great question. Thank you for having me, Hannah. I really appreciate it.

When I think about some of the lessons I've learned over the past years, one of them I learned in education was really focusing on teachers and students. What are we doing? How do we track that has a positive, measurable outcome for those people?

And it's the same thing in multifamily. I'm really proud to say whenever we're talking about a feature, we always pause and go, "Hold on. Is this a positive experience for the renter? Does it help them? Is it a positive experience for a leasing agent? How does it make them more efficient?"

That's one thing I love, that mission that we're actually trying to help real people, not just make money. That's carried over.

On the technical side, I've done product management for 15 years now, more on the technical side than the platform side. Sometimes if you lose sight of, hold on, someone's going to be using what we're doing at the end of the day, and it needs to be simple, you create these really complex flows. You create these really complex UXs.

That's another thing I've carried through. AI is very interesting and new in that it's working on behalf of you, and you need some visibility to know what it's doing and if it's doing the right things. That's another big thing I've carried over: I need a simple UX that shows you all these really complicated things happening behind the scenes, but I need to explain it in a way that it doesn't hurt your face.

Some of the new challenges, I would say this one's pretty common. Just you hear anyone talk about AI. I was really lucky. I got in early with the LLM wave. I've been doing this about four or five years with AI, and it's such a new, quickly evolving thing that I have five newsletters I subscribe to every day, and I still feel like I'm behind.

The new challenge is a technology and a landscape that's ever-changing, and every time you think you're getting it, there's something new. What I've learned there is this is why I lean on my team.

4:49 AI newsletters to read every day

Hannah O'Leary: Two things I wrote down when you were talking. One is auditability. If you can't see what a solution is doing, how do you start to stay compliant? How do you think about governance?

I love that comes up in multifamily. It's something that feels like it had to be built into the bones from the very beginning, just with fair housing and fee transparency.

The other thing, would love to hear names, or maybe you can send these to me after and we'll include them in the show notes, but just five AI newsletter resources that we can share out for other people who are curious about AI.

Hannah O'Leary: For other people who are curious about AI.

CJ Donio: Yeah, I'm a big fan of… I'll give you my top two. I love The Rundown AI. It's a great newsletter. It's free. They do have a premium tier. And then AI Secret is another one I'm really fond of. And then I'll add a bonus one, which is Lenny's newsletter.

Fun fact on that one, they do this special every year where you pay something like 200 bucks or something like that. You get access to 20 or 30 tools. It's really cool because not only are you getting a newsletter with podcasts embedded, but you also get access to new AI tools you may not have heard of.

5:57 Introducing Funnel's new AI conversational assistant

Hannah O'Leary: That's cool. Thank you.

Thanks for the insider scoop of how we can all learn more about something that's moving so fast. Because I think, like you said, it's very easy to feel like a beginner because there's so much new information coming at everybody.

Rumor has it, speaking of new things, that you're working on something pretty cool here at Funnel.

For those that are listening, think of this new functionality as a personal AI assistant for the teams that are using Funnel day-to-day. It's an at-the-ready support team. It's a business analyst that can work alongside teams. It's embedded both directly within Funnel as well as the LLMs that teams use every day.

That's a really long intro, but what has you most excited about this, and what have you been hearing from clients about how they'd actually use it and where it helps them the most?

CJ Donio: This is really cool because there's this problem for the last four or five years everyone's trying to solve, which is how do I get my own personal assistant that knows me, knows my workflows, knows the different tools I use, and it could be my assistant I delegate work to.

This is almost our take on something like an OpenClaw, a Hermes, where you allow it access to Funnel, and then you're able to say, "Hey, I want you to do this stuff to help me in my job."

What I'm most excited about is, I go around often to customers on their onsites, and every once in a while we'll actually go and we'll talk to leasing agents. It's a hard job. It's a hard time in the economy as well, and it makes it a lot harder.

The way I'm approaching this that I'm excited about is, if we do it right, this is a giant efficiency boost. But I also want to make sure it's super clear we don't want to just completely remove the human. The way we're talking about this, the way we're designing it is very much like we give you a superpower, but there's always that human in the loop.

Some of the stuff we're talking about that's really exciting is you basically can query data. You could say, "Hey, how many people didn't show up for a tour today?" And it's going to respond with that data, and then you can actually act on that data. Maybe you want to assign a task to your team in CRM. Maybe you want to send a communication. But this allows you to almost be the coach and tell your players where to be on the field, and we're just getting started.

I'm super excited to see how that evolves to the point where you basically can delegate a bunch of things that would stress you out, and now you can focus on building relationships and following up with renters one-on-one where your time is most valuable.

Hannah O'Leary: What's one thing that's flying under the radar with this first release, and what are you excited to expand and grow after the initial launch of this?

CJ Donio: Oh, yeah. Okay. We have multiple phases to this functionality. We start off with you being able to query data, to act on that data. We'll eventually allow you to reference Funnel documentation. We help you do your job.

I'm most excited about what we call the prioritize phase. We have so much information, so much tooling at our disposal being within CRM, our flagship product. I would love to get to the point where you just say, "Where should I focus? Where are the fires today? Where's my time best used?" And we help you attack the problems that are the biggest.

That I think is going to be a game changer for people, because from what I've heard, sometimes it feels like things are coming at you from all angles. But we can take a data-driven approach to get the most bang for your buck focusing here.

Hannah O'Leary: That's cool. It also feels like it's letting people be more nimble so that they can make changes in what they need to do tomorrow based on indicators they're seeing today, whether that's missed tours or other pieces of information that they could query, so that they're following up with those renters at a time that actually makes sense and we're not looking at lagging indicators of, "Ooh, there's a problem. Now we should fix it." You can get ahead of it a little bit more.

Is that how clients are seeing this?

CJ Donio: Yep. Oh yeah, it's real time. I'll give you a very specific example. I was meeting with a customer about student housing. You basically have two, three months where you are very heavily signing leases to get everybody set up for that term or that quarter. They need to be able, at a moment's notice, to be like, "Has anything changed? Do I have to focus anywhere else?"

It's going to be a big benefit to them because sometimes they don't get back to something for hours and they've missed some opportunities.

10:23 Impact for multifamily data connectors in your preferred LLM

Hannah O'Leary: How does this release connect to the bigger picture? Giving execs their own way to query, I know you mentioned on tours, but also through other tools that they already use.

CJ Donio: Absolutely. We're starting with leasing agents in CRM. We're actually at the same time building a feature for an MCP connector for executives, for data teams.

Basically, that's within an LLM interface, something like Claude or ChatGPT or Gemini. Within that interface, you being able to interact with Funnel and getting information or doing things.

We're actually building that right now for conversational BI. You're able to ask questions about our data. It basically hits a semantic layer, does a SQL query, returns that data, and then you can also say, "Hey, turn this into a table," or, "I want to see this in a column chart."

You can actually transform it into a visualization. And our bet here is, first of all, we're hearing a lot of admin teams that are working outside of Funnel CRM. We need to give them a way to talk to CRM, to interact with that data.

But also we're hearing our in-app reporting's great, but I want to carve it up in a different way for this board meeting, or maybe you're doing an all-hands and you want to slightly tweak it. You can do that all within these chat LLM interfaces.

And then the last thing I'll mention, we're hearing a lot of excitement because customers themselves are building their own MCP connectors, or their other partners are building MCP connectors. So the potential there, I could set up a connection to my PMS, to my company, to Funnel. I could pull in data comprehensively and get this view that I have not been able to get without a ton of legwork in the past.

12:02 What is Model Context Protocol and what that means for multifamily technology

Hannah O'Leary: Okay, I want to break that down just a little bit. There are a lot of acronyms we threw out. Let's define them, how they're important.

BI tools have been around for a long time. I think people know that's business intelligence and an ability to curate and consolidate all of the data that are in the different solutions that people are using and better understand their business would be my 30-second spiel.

Can you talk about how MCPs play into what they are and how they play into this new world of LLMs and being able to pull in data from other places when traditionally in multifamily people might have thought the only way to do that was through a BI tool? I guess would be how maybe we should think through this.

Are there other areas that I'm missing that we should talk about?

CJ Donio: Yeah. When we talk about MCP, it stands for Model Context Protocol. To get nerdy for two seconds, Anthropic came up with this standard, and the standard is basically how to allow AI agents to do tool calls, how to interact with things outside of that chat interface.

I've heard people describe it as an API for APIs. To get non-technical, what this means is you basically want your AI agent to have context from your PMS, from your partner systems, from Funnel. MCP is basically that way to integrate, and the best part, it's pretty lightweight. It's not like an API integration that will change all the time and things will break.

This spec is very fluid to pull data and allow the AI agent to speak to it. The great thing is everybody's using it in the market right now, whether you're Microsoft, doing MCP connectors to their systems, or even companies like Canva or Monday.com have their own MCP connectors that allow your AI agent to interact with their applications.

Hannah O'Leary: The other thing that I've heard, and maybe you can help explain this more too, is that MCP comes along with separate sets of permissions based on what either that user or that company should be able to access. Think of it like a toggle switch where it's, does this person need read/write access? Yes, no. Or does this person need to be able to change these things in a more holistic way, portfolio-wide? Yes, no.

It allows you to set things instead of holistically across the platform like maybe an API would. It allows you to set it mostly on an individual or team level.

CJ Donio: Exactly. That's exactly right. We basically use your same login authentication for your MCP connector. The first time you set it up, you log in and we go, "Oh, you can access these three communities."

So yeah, we only give you the access that you have in Funnel. The great news too in Funnel is we have very robust and granular permissions, so you get to decide. You get to set up someone that has all company-level data, or you can set up someone that's, "I only want you to worry about this individual community."

15:01 How multifamily leaders execute work outside of their tech stack with MCP connectors?

Hannah O'Leary: Thanks for helping break that down and break down how things like that are changing and how they influence the tools that we're able to build and give our partners better information.

When we talk about Funnel being the logical orchestration platform for all of your renter-facing workflows, how do we think about operators leveraging this to execute work outside of Funnel?

And I just want to also back up. There's another term that's floating around the industry, orchestration. I think when we talk about orchestration, we mean orchestrating complex workflows both for AI and teams across solutions across the portfolio. Let's frame it with that in mind.

How are we able to help operators leverage this to execute work outside of Funnel?

CJ Donio: Going back to one of the first questions, this is another example of I've done this across multiple industries, and really the thing that resounds the most is let's enable you to bring your own app stack.

What I mean by that when I say app stack, you have a bunch of partners you have agreements with today. It's not ideal when you sign up for something new and you're like, "Oh, they don't integrate with my partner. I'm going to have to go with one of the ones they do." That's not easy. That's very painful. You might love that particular partner.

Where we're going is we literally want to build this robust integration platform that allows partners to integrate with us, so you can bring whatever partner you want and use them within Funnel.

That might look like our AI agent, like we were talking about before, having an MCP connection, being able to interact with your system, answer questions from your data. It might look like that, or it might look like you want to integrate into our CRM product and have information show up on the guest card or in our ResApp product and show up as a custom task.

We basically want to enable a bunch of partners in the multifamily ecosystem to come together so that you have the optionality, you have the flexibility to use the different tools that you're fond of.

16:59 Getting leasing team members up to speed quickly

Hannah O'Leary: Yeah, truly allow people to run their business how they know works best for their portfolio.

Historically, one of the challenges in multifamily has been getting new leasing team members up to speed quickly. Lots of new processes to learn, new workflows, new software. How does this new functionality help team members ramp quickly?

CJ Donio: Yeah. We were talking about our conversational AI assistant earlier. We absolutely want to support a feature where you can just ask: How do I use Funnel? How does this work?

We do actually want to leverage a lot of our internal KBs for that. We also want you to bring your own processes. We want you to be able to say: Here's all my SOPs for this particular area. Can you teach your AI assistant about this, and then literally coach my team?

We are thinking about it in that way. Also want to talk about a couple other dimensions to this. We have this really cool product called Insights, and basically the way it works, conversations flow into our ecosystem. We analyze that conversation with AI, we extract the valuable tidbits.

The other cool part of that is we can actually listen to leasing agent calls and we can give you coaching. We can say, "Hey, you didn't say this. Remember, you're supposed to say thank you, have a nice day at the end," for example.

We can actually build custom scorecards that allow you to actually develop as a professional and get real-time feedback on your interactions with renters. I'm super excited for that.

And the last thing I'll say here too is the other thing we're talking a lot about is we love to imagine the renter journey and we love to think about, okay, where can we meaningfully put automation, AI interactions to help you?

The other cool thing is we're building a lot of our systems in a way where if you have a newer team, you can add more automation to give them time to ramp up. If they're doing really well, you want to focus more on that human connection, you could reel back automation and AI and give your team more room for those one-to-one connections.

Super excited about where we're going. Not only are we going to help you develop by giving you some coaching, but we can also have your back if you need some time to get people up to speed. And then remember, within the app we're able to answer questions about your SOPs and how Funnel works to get you to that next level.

19:03 What's next? Generative UI

Hannah O'Leary: That's really cool. Is there anything else that I haven't asked you that feels important to share about this AI conversational assistant and how it will help teams, help operators? Anything else?

CJ Donio: I'll say this. Again, AI is ever-evolving, so I'm going to use a really AI-y term called generative UI. The premise of this idea is you have a chat interface and you say, "Here's what I need," and literally a unique UI gets generated that solves your problem.

We're thinking a lot about that. We're thinking a lot about instead of relying on static flows, "Hey, click these three places to get where you need to go," there could be a future where we actually start to serve up what you need in that situation.

I know that might fry some people's brains, but that's really where we want to get. You explain your job to be done to the AI agent, and it's like, "I know exactly what to do," and it guides you there.

Hannah O'Leary: Yeah, and that seems really powerful when you think about Funnel being the home of orchestration, also being able to pull all of that in a more generative way. That's cool. Thanks for sharing the crystal ball a little bit with us.

Okay, switching gears. Part of what makes Funnel really great is also just the people. Obviously we have a great product. Obviously we love working with our clients, but the pink piece of Funnel is near and dear to all of our hearts.

So who is CJ outside of work? What do you get into when you're off the clock, and how does that influence how you show up here at Funnel?

CJ Donio: That's a good question.

Hannah O'Leary: I was told to just tell you to go drum.

CJ Donio: Yeah, fun story there. I actually did once. People finally bullied me enough and I did it, and the web conference muted it, filtered it out. But no, I love music obviously. I have a two-year-old, so she keeps me pretty busy, and she actually loves drums. We'll come down here, she'll just hit on them. She just smiles ear to ear. It's really cute.

What I love about Funnel is I really do enjoy the people I work with, and they know a lot about me, and they know a lot about my daughter and my hobbies and stuff like that, and we bond.

Honestly, my favorite part is people like you I get to work with every single day, and I just feel lucky to enjoy the people that I work with, and to feel like I have 200-plus friends all in one company.

Hannah O'Leary: Well said. I think that's a shared sentiment across. Thanks for being here. Thanks for joining Inside the Funnel.