The Future of AI in Project Management with Forecast, an Accelo Company

Tim Fisher (00:01.19) Hello and welcome to the future of AI and project management, a series where we go beyond the pitch deck and dig into how AI actually is changing the way teams plan, deliver and manage work. Today we’re featuring Forecast PSA, an Accelo company. If you’re running a professional services team and you’ve been wondering where AI actually delivers value, not theoretically, but in the tools you use every single day, then this is the conversation to be in. We’ll hear their perspective on what’s broken in project delivery, see their AI features in a couple of really cool live demos and talk about where all this is heading. So drop your questions in the chat as we go and we’ll tackle as many as we can towards the end of the call. My name is Tim Fisher. I am VP of AI at the Digital Project Manager. And my job is to figure out where AI actually works and where it’s still just marketing slides. So I’m gonna be asking the questions that your team would ask. Joining us today is Joe DiPaolo, CEO of Accelo, the company behind Forcast PSA. Joe spent his entire career helping professional services teams build more predictable and scalable delivery organizations. And today he’ll share how AI is shaping that future. Joe, great to have you here.

Joe DiPaulo (01:07.502) Awesome to be here, Tim. Thank you for having me. I’m excited about the topic and how this is shaping our professional services world.

Tim Fisher (01:10.554)
Of

Tim Fisher (01:16.088) Very, cool. Joe, just give us like the 30 second version of Forecast. Like what does it do and who’s it built for?

Joe DiPaulo (01:23.822) Sure, so Forecast is a modern PSA or professional service automation platform. It’s designed for professional services firms that are typically growth oriented and they’re working to scale their businesses. So I think that’s the most important part. We focus on functions from project planning, capacity and forecast, delivery execution, revenue operations. So our mission is really to give our customers visibility into their data, predictive insights into what’s taking place with their business and how they’re delivering, and help them manage their business as effectively as possible.

Tim Fisher (01:58.652) Awesome, very cool. So, Forecast recently joined forces with Accelo. What does that combination unlock that neither platform really had on its own before?

Joe DiPaulo (02:08.674) Yeah, the goal with the combination of the two organizations and platforms is to bring a true end-to-end operating system for professional services organizations. Forecast has been really exceptional at AI-driven project planning, resource forecasting, demand planning from a resource management perspective, and providing a lot of good forward-looking visibility. Acello has a lot of strengths in CRM capabilities, client workflow management, financial workflows. So together that connection kind of covers the whole pipeline from our perspective of a professional services organization and acts as a key system of record.

Tim Fisher (02:46.748) Awesome, that makes sense. So a nearly comical question. Why is AI the thing you guys are really being into hardest right now? Like what’s the strategic bet here for you?

Joe DiPaulo (03:01.602) Yeah, I don’t think anybody’s talking about this, are they? We’re leaning into AI because for our customers that are service-oriented businesses, they’re operating in a very margin-constrained, complex environment. There’s a lot of factors that go into delivering for their customers. And historically, these businesses have scaled by adding people.

Tim Fisher (03:04.444) No, not at all. No, never comes up.

Joe DiPaulo (03:27.72) So the bet with AI and more intelligence and automation is that these businesses can leverage that as a tool to help them improve their performance, improve the customer experience, and ultimately scale and be a more profitable business. know, AI in our perspective allows leaders to move from what’s happened historically and being somewhat reactive to helping them really have a much more proactive type of business. Again, I’ll probably say this a lot today, but it goes back to allowing our customers to scale and ultimately scale without adding headcount and allowing teams at our customers to really focus on the value add of their services and build relationships with customers and make sure that they can continue to grow to optimize their delivery process. That’s why our focus is there and it’s ultimately because our customers want us to help them grow their businesses.

Tim Fisher (04:30.587) That make sense. So before we jump into the demos, I want to get your take on just what you’re seeing across the industry right now. So you’re talking to professional services firms every day, like agencies, consultancies, IT service shops, the lot of them. What’s actually changing in how work gets delivered right now?

Joe DiPaulo (04:50.668) Yeah, I think our customers are evolving from experimenting, I would say, with AI over the last year or two to really trying to leverage it more effectively as a resource in their organization, how they deliver work. They’re looking at, you know, gentifying their business and using that to help them deliver for their customers. And I think everybody’s at a different stage in that journey, but they are working to produce repeatable results and optimizing those results through AI and I think that’s the evolution that we’re seeing right now.

Tim Fisher (05:33.757) Yeah, I think we’ve all seen the news articles recently about AI not performing for organizations or not living up to the hype or the promise. I think there’s definitely a focus right now in real practical outcomes. And it definitely feels like what you guys are after here. You’re doing a little digging. And your platform data shows that teams report utilization about like 75 % compared to 40%, which is industry average. Like what do you think drives that huge gap?

Joe DiPaulo (06:10.029) I think it’s, you know, when you’re able to operate more consistently and when you’re able to use tooling to help you be more predictable in what you’re doing, you can be much more proactive. And I think one of the things that we try to do in the PSA system is to give our customers insight into work that’s coming down the pipeline from their sales organizations, drive repeatable, consistent processes, and allow them to see who’s doing what and when. And the more that they can predict and be proactive about those things, the better they can operate. So when they can plan ahead for work that’s coming one, two, three months down the road, or when they can see exactly what everybody’s working on, they’re able to not overwork individuals, but to leverage them to their best and highest potential and capacity. So the more that they can leverage those operational capabilities, the better off that they can run their business and we see that translating into better utilization results.

Tim Fisher (07:20.826) Yeah, very good. Very related. You look across these 2000 plus professional services firms that you guys work with and what are all the biggest gaps that you see? Whether that’s margin leakage or this famous thing that I think is still happening today, which is resource management off of spreadsheets and all the spreadsheets. Just lay visibility into problems. What does that all look like to you right now?

Joe DiPaulo (07:47.437) Yeah, I mean, there’s a couple different topics to unpack there, but in terms of margin leakage in professional services, you certainly, that’s something that our customers tend to struggle with in terms of, you know, it’s never something that happens all at once, right? It’s something that happens, you know, bit by bit as a project or an engagement with a customer goes over time. They’ll see, you know, issues that they might have in terms of how they assigned resources or which resources they’re leveraging. Do they the right people? Am I over utilizing, you know, a certain skill set or a certain high performer?

Tim Fisher (08:15.676) Yeah.

Joe DiPaulo (08:22.58) They’ll often see delayed visibility into risks in their work and projects that are going adrift. Things like that are what we try to solve. I think another thing that plays into margin is scope creep. So we’re really leveraging the system to help them detect and predict when they’ve got challenges like that. And ultimately that results in margin improvements. I mean, again, there’s lots of places in the lifecycle that this can happen, but the more that we can alert and help customers proactively predict when those things are taking place, the better that they can respond to that. In terms of working off spreadsheets, I think that’s common for smaller organizations. And lots of times people say, I’m more agile this way. I can be more responsive. But I think that’s a bit of a fallacy. And obviously, we’ve got a tool that helps solve that problem. As soon as businesses grow in their complexity or the number of projects that they’re doing, the different types of pricing models they structure, how their engagements are structured, the more teams that they have, that becomes more challenging. You lose some of the accuracy and the real-time controls. You start to break patterns or break processes that will impact margin and kind of your discipline around margin. know, in organizations that are really leveraging data and, you like I talked about before, in work from their CRM and analyzing that and using their looking ahead down the road to figure out how do I need to staff this, what are the people I need. They’re able to really see the value outside of just managing the process in a spreadsheet and get a lot more predictive and hopefully run their business much more effectively and smoothly.

Tim Fisher (10:24.688) Fantastic. So Forecast was built AI native from day one. I don’t think a lot of PSA tools can say that. That’s not something I’ve actually heard before. That’s impressive. Like given what we just talked about, know, like margin pressure, scaling, visibility, all that sort of stuff, like what does this actually mean in practice for just, you know, for like how your product tackles these problems differently?

Joe DiPaulo (10:50.624) Sure, yeah, that’s right. We’ve had, the system is really founded around the concept of leveraging machine learning models to help teams do better staffing and resource planning and ever since the beginning it’s evolved and grown. The goal is to have the AI act as a guide for professional service organizations. So helping them address inefficiencies, helping them get past scale ceilings or scale challenges that they have. As we look at it, as firms can get more predictive, they can avoid risks, they can drive a better customer experience, they can leverage their resources more effectively and optimize how they’re leveraging their resources so they can hopefully keep their teams happy and high performers from being overutilized and burned out. So the AI is going to help them kind of identify those concerns and risks, drive efficiencies. One of the other things that we hear from customers is as I’m managing projects or engagements across different parts of my business or different departments, they tend to be able to see data better across their entire organization. So if they need to leverage resources cross-functionally or make decisions that are across those barriers, they tend to be able to do that effectively. So AI is certainly helping with that. And then as you’ve hinted at, as we look forward, AI will certainly get more and more central in how professional services firms actually do work and deliver outputs and that’s the exciting part about what’s taking place today and again going back to how businesses scale and how AI will allow them to do that more effectively.

Tim Fisher (12:39.612) You mentioned machine learning, which I feel compelled to bring up because we all use the term AI and even use the term like large language model. But just dropping that phrase tells me that you guys have been working on this before Chachibet dropped into the world. And this is more foundational to how you’ve been thinking. And I think that’s very cool.

Joe DiPaulo (13:02.242) Yep, I agree. It’s exciting. I’m glad that we had that history and that started years back. But we’re certainly continuing to invest and take advantage of all of the newer and newer technology that’s coming out to help our customers.

Tim Fisher (13:17.712) Definitely. OK, let’s see this in action. Joe, can you walk us through some use cases of Forecast’s AI capabilities?

Joe DiPaulo (13:27.018) Sure, I want to go through three use cases today and let me work to share my screen here.

Joe DiPaulo (13:39.823) So I’m going to talk through three use cases, like I said. I’m going to do two in our application, and these will highlight some of the machine learning models that we’re leveraging to help our customers. The first one I want to show is diving into a project and showing in the system where…

Tim Fisher (13:42.652) Okay.

Joe DiPaulo (14:02.966) …we might predict that a project or a portfolio project is running into an issue or how we might help a project manager or a leader at an organization do that. So I’ve logged into my dashboard here as a specific project management type resource. I’m going to go into a specific project that’s actively running. And as I look at this, I’ve pulled up what we call Nova Insights. That’s one of the names that we use

Tim Fisher (14:13.852) Okay.

Joe DiPaulo (14:32.88) …for our AI tool and this is constantly analyzing the data around all of the activity in the system. So all of the projects, all of the efforts that are getting completed and delivered to customers, hours that are being tracked, work that’s being done and invoiced. And in this specific scenario, you can see it’s giving me some insight into some very specific details about this project. So, you know, it’s looking at things like task performance and basically telling me that tasks are generally being completed along how they were estimated. However, it looks like some of my resourcing on this project is really under, you know, not meeting the timelines or not as available as I expected when I started the project and when I built out the estimates. And based on that, likely it’s saying, well, you’ve got some risks from a budgeting perspective and you’ve got some risks from when the project will be completed. And I can drill into these and see a bit more detail. So if I look at the budget information, you know, you can tell I’m relatively early in this project in March, but the system is already projecting a variance in…

Tim Fisher (15:19.91) Yeah.

Joe DiPaulo (15:47.461) …in terms of my margin and where I’d hoped to be or what parameters I had set up from a margin achievement perspective. So it’s suggesting there’s risk in that regard. It’s also projecting that the end date is going to be past… past what was initially projected. And this is probably an extreme example. This is demonstration data. So it may not be perfect, but it gives a good indication of what’s taking place here. In this case, the project was booked to end in May, and the system’s saying, based on the amount of work that’s done towards this, it actually projects it not being done until August. So from that, the project manager or the person running this project can…

Tim Fisher (16:17.629) Sure.

Joe DiPaulo (16:35.436) …take action. The system also has what we call assists where it can suggest how you might be able to resolve some of those challenges to get the project back on course in this case. So, you in this scenario, it’s estimating that we’re light on these roles for a certain number of hours and gives a straightforward way for the person running the project to adjust and consider who’s available for the project. And from this screen you can actually create what we call placeholders in the system and those create demand for specific resources that then I can leverage to assign. So that was the first use case that I want to show. It’s a good example of how the system can predict risk and concerns with a project or a group of projects and you can see how somebody can leverage that to quickly adjust or review or dive into a project more deeply.

Tim Fisher (17:33.618) Very cool. A question for you. How much historical data does the AI need before those recommendations become useful? And the other side of that question, guess, is there a kind cold start problem with just an organization not having enough data in your system to make really good capacity planning answers?

Joe DiPaulo (18:00.866) Yep. Yeah, it’s a good question. Models definitely need a bit of information to learn from and to be able to leverage. However, during our implementation process, our best practice is to load a year or two of historical data. So if you can put in resource information, historical project performance, you can shorten that window. So that would be available almost immediately. And part of our implementation team’s process is to work with new customers to bring data into the system to help them get value out of that from day one.

Tim Fisher (18:14.854) Okay.

Joe DiPaulo (18:33.122) So they’ll work consultatively with the new customers on that effort.

Tim Fisher (18:33.893) Excellent.

Tim Fisher (18:38.653) Cool. Cool. All right.

Joe DiPaulo (18:40.938) So maybe continuing in the journey here, Tim, I want to talk a little bit about scheduling and…

Tim Fisher (18:47.741) Okay.

Joe DiPaulo (18:48.502) I’ll go into the second use case where I want to focus a bit more on capacity planning, resource planning, and how resources can get assigned. this goes to your question about managing in spreadsheets versus managing in a tool. And so this is a very high-level picture at this point in terms of capacity by month. I can do some things like drilling down to different weeks or different time periods and go further out into the quarters. But this is showing a resource…

Tim Fisher (19:02.598) Yeah.

Tim Fisher (19:13.415) Very cool. Yeah.

Joe DiPaulo (19:18.346) …manager or somebody who has to do some planning with resources and assign them to work, what upcoming work that we have, and this can be filtered in different ways by specific projects or customers or however they’d like to do that, but if I drill into this example, I’m looking at work that I’ve got in flight. In this case, in the month of March, I’ve got, you know, hours assigned to people that are actively working on those, and then I’ve got this concept of placeholders, which is what I

Tim Fisher (19:33.511) Yeah.

Joe DiPaulo (19:48.129) …referred to previously in the assist option. And you can see specifically for project managers, I have different work that is out there that I need to assign. When I go through and assign this work, one of the things that the system will do is, it shows that I’ve got eight people matched here. So it’ll actually make recommendations for that work based on things like the skill set, the role that the person plays, the rates that they have assigned to them, and the different attributes from their profile. So this should make it relatively easy for a resource manager. If I dive into that even a bit further and look at this detail, the system will actually show what it thinks are the top three resources for this specific work. So I’ve got this project management, 260 hours, I can see that’s going from March through May. And I can look at different resources here. can see what do they have assigned to them. I can actually drag and drop this and move this around to see how that affects their capacity for different hours. And if I want to assign that work to them, I can. But the point here is that the system is taking some of the work out of…figuring out who’s available, what’s the right skill set, how do I assign this work. So that’s all done with those models that I talked about and making those recommendations and making it easy for different people to do this. And I said they can do that at different levels depending on how each organization works to assign resourcing.

Tim Fisher (21:29.309) Very cool. And maybe you’ll get to this, just like a thought as I’m seeing this, an action, like when the AI makes a recommendation for like a role, what signals that? Like what powers that? Like the skill match availability, past project performance, like what plays into those like smart recommendations?

Joe DiPaulo (21:50.007) Yeah, it’s exactly those things that you’re talking about. mean, I think the first most obvious one is availability, right? If I’ve got work that needs to be done tomorrow, who’s available tomorrow? But then it’s their role, their skill set. And we take in a lot of data in the system in terms of past project performance, what kind of tasks that have they worked on, how well have they achieved those. And it’s an interesting part that the system will learn from. Sometimes a more junior person will do more senior tasks, and that’s part of their evolution and the system is aware of that and then can take advantage of that in these recommendations. but those are the things that come into it.

Tim Fisher (22:27.869) I haven’t seen a spreadsheet do things like this. So it’s very cool.

Joe DiPaulo (22:31.342) I agree with you. The last use case that I want to cover, Tim, if it’s a good point to move on, is I want to show a little bit more of we’re leveraging MCP to allow customers to interact with the data in our system via any kind of third party LLM that they want or through other systems so that they can build their own AI workflow. So I’m actually going to pull up Claude and walk through some of that if that makes sense.

Tim Fisher (23:06.779) Yes, it does. And while you’re doing that, just in case there’s some less technical folks listening in right now, MCP stands for Model Context Protocol. anyone who has done any sort of integrating an AI system in with any of the major like chat platforms, like it looks like we’re going to take a look at Claude here, has probably used or knows what MCP is. But for those of you who are just technical enough to know some of the underpinnings, it’s a bit like an API for large language models and AI models to interact with each other and other pieces of data, other systems. That’s what we’re talking about right here. And so I’m excited to see this in action. That’s very cool that you guys have incorporated that into these tools.

Joe DiPaulo (23:54.723) Yeah, thanks for explaining that, Tim. That was helpful. Yeah, and so, and maybe to layer onto that, I’m using Claude, as you said, but our customers are using various tools to…for their agent and AI usage and you can connect any one of those that you want to into this. So customers can ultimately leverage their own LLMs to get the most value out of not just forecasts, but a group of systems if they want to or to design workflows. And that’s one of the things we’ve heard from a lot of our customers. So this is a beta feature that we’ve been working to roll out. But I can ask things like, who’s available next week? And as I go through that process, it’ll connect to…via MCP and it’ll connect to the data for a given customer and return that back to them. So you’ll see how this happens here. One other thing to call out from an MCP perspective is it is a secure protocol, so customers have to authenticate to use it. They’re only getting the data that’s from their specific account. This is not data that we’re sharing across customers. It’s locked down.

Joe DiPaulo (25:09.614) So that’s part of the protocol. So you can tell it’s thinking here, but it’s looking for availability next week. It should provide us some data back. So again, this could be a resource manager. It could be somebody who’s more in a leadership role looking for reporting information that they want to get out of the system. Could be looking for project status, those kinds of things.

Tim Fisher (25:14.749) Yeah. This is so cool.

Joe DiPaulo (25:37.742) Hopefully Claude returns here in a second.

Tim Fisher (25:39.998) While it’s doing that, maybe a quick comment. When I see tools like this, when I see interactions and workflows put together really well like this. I think about the meeting killer aspect of all of it and like the questions leadership has, I’ve been in these roles where there’s just like complicated systems and complicated products and be able to go to like a user interface and have a user experience that we’re all now very familiar with and get deep contextualized answers from inside your business is incredibly valuable.

Joe DiPaulo (25:55.149) Yep, yep, it’s great. So this is obvious demo data here again, but for next week I can see I’ve got lots of hours available, 55 people with capacity, you know, probably on this list somewhere and says I’ve got 40 hours free which is inaccurate but you get the idea right it’s going to show me information about about people on my organization how much time they have who’s available to work those kinds of things and just one more quick example here and then we’re limited on time but if I if I’m interested in more of a you know executive type use case where I want to see reporting type of data or kind of what you were talking about in terms of the meeting killer you know I can say something like…what is all the log time billable and non-billable? And it might take a minute to run that and get results back, but this is more of a reporting style scenario where instead of running data into a spreadsheet or a dashboard, I can query very specifically what I want to look for and the system will go through that process and pull this back. And again, this could be from any LLM that a customer is using to get data out of the system. So it’s thinking here for a minute, but it looks like it’s pulling back lots of different time entries. It doesn’t look like in this demo scenario that my utilization is very good, but you get the idea of hours logged, how things are being billed, non-billable hours.

Tim Fisher (27:55.709) Yes.

Joe DiPaulo (28:01.198) …looks like it’s still summarizing some of the data and building some information. But it’s a good example of that MCP use case and allowing customers to connect to data in their system of operations to then answer questions and provide more meaningful insight into their business.

Tim Fisher (28:20.741) As funny as you are describing this, can hear in the way you’re describing it a little bit of hesitation around like how long these things are taking. But if we zoom out for a second and think about the work going on behind the scenes and the amount of time and effort it would take to pull all this together, this is incredibly speedy. This is really, really, really cool.

Joe DiPaulo (28:38.702) Yeah, for sure. mean, we’re talking about 23,000 hours and lots of data here. So it’s pretty quickly going through and returning both contextual information and detailed information about the project. And again, this is very much kind of demo kind of data, but you get the picture right in terms of what it returns. And in a matter of 20 or 30 seconds, you’re getting a really deep overview of this specific client that I’ve asked about. So those were the use cases I wanted to cover. Hopefully that was helpful.

Tim Fisher (29:10.661) Amazing. Great. Definitely, definitely helpful. Okay, so we just saw what AI can do today, which is, I think, really, really impressive. But let’s take a few minutes and look a little bit forward. So the PSA market is projected to nearly double by 2030, which is wild to think about. And obviously, AI tools and project management are growing even faster than that. So Joe, do you think this goes from here?

Joe DiPaulo (29:43.611) I hinted at this at the beginning, I’m really excited about how AI is enabling professional services business to leverage agents to help deliver for their clients more effectively. We’re going to see organizations use agents to… to deliver work and take some of that off of humans. And that’s going to mean quicker time to value for their customers. It’s going to mean a better customer experience. And it’s going to mean better profit margins for the users of our system. So I think that’s an exciting proposition.

Tim Fisher (30:15.613) Yeah. Definitely. you can share what you like or at whatever level. So I guess without giving us your full roadmap, what’s the capability gap that you are most excited about closing with AI and how does that change what your customers are going to be able to do?

Joe DiPaulo (30:44.47) Yeah, it’s building on that theme that I was just talking about. I we certainly hear tools like Claude and others like it helping software development teams, right? And helping to write code and do that more efficiently. we want to be part of that ecosystem and help our customers kind of identify their workforce and allow them to orchestrate and connect into their, with their operational systems into those agents to most effectively, you know, run their business, see information in context of their business and workflows, but take advantage of the capabilities that agents are offering to make their work more robust and deliver it more quickly. So we’re certainly believers in that. We’re doing that within our own company and we can see how that can affect and improve our customers’ businesses quite dramatically.

Tim Fisher (31:48.166) Awesome. So I like to ask this question a lot. like we are three years from now, three or four years from now. What does day to day look like to you for a project manager? Like how does that role evolve? Is there even a project manager? Do you see AI taking that role over? Like how do you see the future from your vantage point?

Joe DiPaulo (32:10.158) Yeah, I mean, I think we’re going to see massive shifts in how work’s done. I think that there’s still going to need to be people involved to help organize and think about the outcomes that we’re trying to achieve and really throwing a buzzword around, but really help orchestrate these systems to deliver. So there’s going to have to be a lot of thought process into what are the inputs that I’m putting into these machines? What’s the outcome that I wanted to achieve? There’s still going to be quality controls and things that are going to have to be done. But I think all stuff is taking place faster than the next three years. We’re going to see significant changes and agents acting as workers in organizations. If we’re not seeing it now, it’s certainly going to happen in the next 12 to 36 months and it’s going to get more more adopted by different, very specific, across many industries I’d say. So that’s, yeah, think that’s fascinating.

Tim Fisher (33:13.554) Yeah.

Tim Fisher (33:18.971) Yeah, it very, much is. So, okay, thank you so much for walking through all that with us. Like very, very insightful conversation. So for anybody watching and thinking, okay, I want to see forecasts in our environment with our data. Like what’s the best next step for them to take?

Joe DiPaulo (33:35.343) The best thing they can do is to come to our website, it’s Forecast.App. They can request a demo and a consultant will reach out to them. Usually they’ll want to understand the business that they’re running a little bit so they can tailor demonstration discussion and kind of define how we would show value for them and go through that process with them. that’s what I’d recommend. We’re also, part of this webinar, be publishing an AI readiness assessment so that’s another thing that customers can interact with and engage with and we can certainly reach out and contact them if they’re interested after that as well.

Tim Fisher (34:13.629) Fantastic. Awesome. Okay, so we’re to move into the live Q &A now. So for everyone listening, you haven’t already dropped your questions in the chat, this is your opportunity to ask the forecast team your questions directly. We’ve got some already and I’m going to walk through some of those with you right now. Okay, so Yaron asks, how do you reliably compare time saved on project management issues with and without AI? Is there such a thing as AI rabbit holes?

Joe DiPaulo (34:40.622) Yes, so mean without AI it’s usually done through reporting or dashboards or other mechanisms like that in the system. You can look at it on an individual project basis at a very detailed way or you can look at cross using reporting to traditional kind of BI reporting or dashboards. You know in many cases we’re using a lot of the same data to populate the language models or the AI tooling that we’ve got to provide that. So you should get pretty similar results, know, obviously depending on how you’re prompting agents to get that information back. But the backbone of the data is the same. And I think that’s one of the important parts about using AI in an operational system like ours is the inputs are consistent, the data structure behind it is consistent. You know, there’s not a lot, it’s not going out to query the web to get context for hours being worked. It’s the data within your organization. you know, from our tests and results, we see very similar types of results in those looking for information.

Tim Fisher (35:50.256) Okay, we have two very similar questions from Michael and Eduardo. Will this type of functionality be coming to Accelo? This is in response to, I believe this question came in while you were talking through your roadmap.

Joe DiPaulo (36:03.724) Yep. We’re putting additional AI functionality out in both. Our first focus on the MCP server and MCP workflow is on the forecast side, but ultimately we want to bring it to all of our customers.

Tim Fisher (36:20.744) Very cool. Vivek asks, where is it getting the hours available from? And this was during your final use case, which I believe was the cloud one.

Joe DiPaulo (36:30.222) Yes, so I’m not sure if they can see in our application, you you can. Many of our customers track time from their consultants. They don’t necessarily have to bill that way, but that’s an important part for us to understand costing on tasks and projects and how it compares to estimates. That can also be, you know, entered into the system in other ways, but part of the application is tracking time of individual resources and then, you know, we’re able to leverage that or tracking task completions to estimates and we’re able to track that and then report on it. So that’s part of the core data in the system.

Tim Fisher (36:40.062)
Sure. Very cool. And now some like, would call these like kind of auxiliary questions, just sort of around what we’ve talked about today. A very specific product question. Can you interface with service now or Azure?

Joe DiPaulo (37:19.47) Yes, interface with Xero. That’s a very common, know, accounting and finance systems are very common for us to integrate with. We have standard out-of-the-box integrations with.

Tim Fisher (37:31.44) I’m sorry, I probably mispronounced. I meant Microsoft Azure.

Joe DiPaulo (37:36.258) Got it, got it, got it. Yeah, so through the MCP model, if somebody’s using Azure or Microsoft’s copilot, for example, to pull in data into an LLM in that environment, they could do that.

Tim Fisher (37:43.944) Sure. Yeah. Very cool. Yaron asks, do you have a recommended glossary or index of terms related to today’s topics for further reference?

Joe DiPaulo (38:01.166) That is a good question. don’t know that I have anything offhand, but I’m sure we could get something together if there’s some questions after this and follow up.

Tim Fisher (38:09.873) Great. And sort of related, just as like related information, do you conduct any courses on AI in project management? And if so, any sort of details around that?

Joe DiPaulo (38:23.586) We have not historically done that. We’ve primarily focused on delivering the software and helping our customers take advantage of that more so than training their teams on topics like project management. So that hasn’t been our focus historically.

Tim Fisher (38:29.073) Yeah. Yeah, great. So let’s see, we have another one. Do you think AI can ever replace collaborations and conversations in terms of resource management?

Joe DiPaulo (38:54.231) Yeah.

Tim Fisher (38:54.365) That’s a big one.

Joe DiPaulo (38:59.39) I think it can provide a lot of contextual information and I think it can help make recommendations. I think there’s still always going to be some value in having conversations and understanding face to face with people, what they’re interested in, they feel like their skill set is, those kinds of things. Hopefully it allows for those conversations to be quicker and more value driven versus just saying, hey, are you available in three weeks versus, hey, we’ve got this exciting project that I think fits your skill set well, what do you think? So, hopefully that’s the combination of those two things to make customers more efficient and leverage data in their organization better.

Tim Fisher (39:47.08) Great. One for me. So what makes all of these things you’re talking about, all your tools, the demos that we saw, different from just using ChetGBT or Claude or Copilot with just a dump of data and then asking questions against it?

Joe DiPaulo (40:03.863) Yeah. Yeah, it’s a really good question. obviously, we believe and know our customers are going to use many tools, ChatGPT, Copilot, Cloud, et cetera, et cetera. And we want them to do that. I mean, we’re building use cases to support them bringing data into multiple systems. But the difference from my perspective, and I think this is true of most SaaS applications that are a system of record or part of our customers’ ecosystem and how they run the business, is systems like ours have more context about our customers and their data. I you saw that with the time entry or the hours data, for example. So, you know, a general AI tool kind of is outside of that operational system. So, you know, it can provide context, it can do some analysis, but it doesn’t typically have all of the inputs in real time that are coming from across your organization, whether that’s new customers you’re winning, invoices and things that you’re billing and running, or other data that’s coming into the system. So it’s constantly able to leverage that. So the combination of the two, think, is what’s powerful. And I think that’s the difference. I think that’s the difference for really most SaaS applications that have operational data. I mean that’s where I think the value is, that makes sense to him.

Tim Fisher (41:30.331) Yeah, that totally makes sense. New question from Augustina. From a PM or management perspective, which other areas do you think still have the biggest opportunity for AI driven improvement? And then we touch on this a little bit, but I’m curious on your answer to this one.

Joe DiPaulo (41:47.94) I mean, to me it’s in general any place that a system can thoughtfully help, you know, execute more mundane tasks that are repeatable and can be based on logic. I think we hear a lot in the media about software development or supporting legal context. I think when I look at the traditional professional services life cycle, doing things like helping more automatically generate proposals that are human than reviews before it goes out to a customer. Or another thing that we hear all the time is, help me generate my invoicing that I need to send to a customer, but alert me of exceptions, right? Does the system think that I’m missing hours or missing some context or, you know, like I talked about, identifying risks so that I can overcome that early. I those are the things that these, you know, patterns or, you know, offloading of certain tasks are really valuable that an AI can help with. So, you know, those are good examples in our system workflow.

Tim Fisher (42:54.14) Yeah, you know, you’re, as you’re saying all that out loud, I’m, I’m, I’m sort of realizing that, a lot of the opportunity here is like, you know, when you think of, of operating a really small company, you’re maybe like a solopreneur and you carry all of these things in your own head. Like you just, you know them all, but it’s your business scales. You aren’t as scalable. And so there’s all of these things that you have to weigh into your decision-making that just becomes simply too much for, a single person or a small group. so having a computer that understands language and understands context and everything like that, there’s just so much opportunity in this space, I think, to help people scale without scaling the human workforce that I think you mentioned earlier. So it’s all very, very exciting. Let’s see. We have another one here from Pia. Do you have any?

Joe DiPaulo (43:40.738) Yep, it is exciting.

Tim Fisher (43:48.984) Use cases top of mind where Accelo is being used in European companies, particularly in the tech startup ecosystem?

Joe DiPaulo (43:56.311) Yes, for 100%. I mean, I think that’s one of the things that we’re leaning into most. One of the things that the system does well is we’ve talked a little bit more about more what I call traditional project management in terms of milestones and tasks and those kinds of things. The system also supports a more of an agile type of flow in the system, whether that’s a sprint or a Kanban style type of effort. And we’re certainly leaning into allowing our customers to drive inputs through that process into what’s traditionally been software developers or engineers, whether they’re building something from an agency perspective, like a new website, or managing some technical aspect of SEO, or delivering software for a customer. But we’re leaning into the agents helping software development organizations and doing those things more more agentically you know like we’re seeing with with Claude or with OpenAI so that’s that’s definitely directionally where we’re going.

Tim Fisher (45:12.23)
Awesome. Maybe close to some of final questions here. I think maybe the best one I’ve seen so far. Katarzyna asks, how do you recommend persuading organizational PMO to allow the use of PM tools and software for portfolio management instead of spreadsheets? How can they show the benefits bigger than the risks? And before you answer, I’ll say one of the biggest challenges I continue to run into is like, change management, regardless of the technology or the change, right? It’s like change is hard. And so what would be your recommendation here?

Joe DiPaulo (45:48.847) Yeah, I agree with you on the change management aspect. You know, I think in those cases you have to be able to say what are the pain points that, or understand the pain points that those people are experiencing. For me it’s things like… you know, are the handoffs difficult between my sales and my delivery organization and delivery and finance? And if those things exist today and you’re managing some of those things via a spreadsheet, you know, that tends to be an indication of a challenge or something that hopefully that person would understand and see the value and the return on investment. To your point about change management, you know, it’s usually something that human nature gets involved in that piece and you might have to convince and train and probably have a well thought through plan. You know if you say hey tomorrow we’re gonna start doing this new thing they may not love that but if they’ve got a you know 30, 60, 90 day time horizon and they’re kind of brought through the journey that’s obviously a better approach to driving success.

Tim Fisher (46:56.092) Yeah, I think your last comment is really important. And I think it speaks to maybe the different head spaces a lot of folks and organizations are right now, where you have folks that have been experimenting since day one, and you have folks that are maybe getting into it, and you folks that are maybe actively or passively resisting. And so that’s a gamut of perspectives in especially larger organizations that it’s difficult to manage. But yeah. So, let’s, wrap up with maybe one or two more here. So again, let’s just look a little bit in the future. And I, I, I know I got at this a couple of times already and we even had a question, but I think this is just like, just one of the most interesting things to talk about. like, let’s say you, Joe, as CEO, you can just wave a magic wand and what’s the one single problem. So I’m asking you to actually like, like focus on one thing in project management that you would love to solve next with AI.

Joe DiPaulo (47:56.782) Yeah, that’s a good question. I think it’s taking on mundane tasks that humans have done and helping deliver those things faster and more effectively and allowing the humans involved to build relationships and really optimize their businesses. And I mean, that’s what it is for me from my perspective. And that’s going to vary a little bit based on the type of organization and business you’re running. But I think that’s what’s taking shape very quickly here.

Tim Fisher (48:28.049) No doubt. Yeah.

Tim Fisher (48:32.83) That is the promise of all of this, think, for sure. Yeah. All right. Well, that’s all the time we have for today, Joe. Thank you so very much for joining us. This is fantastic. And to our audience, thanks for all the great questions. And have a great day. And we’ll see you at the next future AI event in a project management session. Thanks, Joe.

Joe DiPaulo (48:52.46) Yeah, thank you so much, Tim. I appreciate it.

Tim Fisher (48:54.589) Awesome.