The Making of Nitro

Yo. Yo. Every company in tech has an AI story. Some of them have a chatbot.

Some of them have a copilot. Some of them stick AI onto an existing feature and hope it secures them another round of funding. Hey. We don't judge.

The product team at Rocketlane here claim that they've built out an entire AI agent equal workforce with Nitro. So to find out more, I spent my last week interrogating the people responsible for building this out. This is the story of Nitro.

So compared to other companies, do you think anybody else in the horizon is doing anything remotely similar to Nitro?

In our space?

No. Nobody? Nobody. Not a single person? No. Not your Here's a Not a not a single person even got, like, agent tech AI going in their PSA platform.

They're trying to tell the story Okay. But I don't think they're succeeding.

So I genuinely tried reaching out to our competitors for comment, but they blocked me on LinkedIn. So yeah.

I would say Nitro is something that changes the way people do work today.

Nitro is our version of bringing that future of work to a professional services team.

Where you're able to connect and bring data from other systems and then do that work with AI right within the platform.

See, software software was always about tracking work. Right? So if you were a CRM, you were tracking a deal. If you were an ATS, you were tracking a candidate. If you were Rocketlane, you were tracking a project, you're tracking utilizations, you're tracking financials. That's generally what it was. For the first time now, software can actually do the work.

So things that used to take people five hours of work, Nitro can just do that in ten minutes now.

The kind of work that a production services consultant would do starts from solutioning, documentation, configuration, testing, again, integration work, and also data transformation, data validation for migrations.

Imagine you have call transcripts from some other platform and or, you know, you have conversations with your customers over email. Upgrading the status in Rocketlane used to involve just going to Rocketlane, opening that page, clicking on a couple of things, then that can take a couple of minutes, and then that adds across the entire month. But then with Nitro, you can just connect the MCPs to the right sources of your data, where Rocketlane becomes the sync. And all people need to do now is just provide a prompt to get the data from the source and then confirm that this is what you wanted to be updated in the platform, which is Rocketlane is the sync. So take the data from the source, put it in the sync, and AI just manages to do this in ten minutes.

Any team that is doing project based services for customers, how do we help them run their business, how do we help them run their projects, and how do we help them execute their projects? The actual service work that they do for their customers, how do we bring AI to do that work as well?

I would break it into back office work and front office work. So back office work is typically work which lets your business keep taking over. Right?

For example, most software in our category have stuck to the definition of the category itself. Professional services automation is about helping people plan the work, staff the work, measure the work, manage the work, maybe communicate the work. And that's where most companies have stopped their ambition.

For example, you're supposed to change fields or mark their task as done. You're writing minutes of meeting. You're sending a project update. You're documenting, you know, what someone someone's requirements were, and there's front office work. When I say front office work, this is work which is actually adding value to the customer.

That's where we felt there was a great opportunity to say, okay. How can we build with agents that help in each of these activities? What if we also bring elements of automating the doing of the work itself?

Translating the requirements to a configuration. How are you moving data from their previous vendor to yours? And that's where front office work comes in. Right?

We have agents for doing the configuration for you, doing the documentation for you, doing the migration for you. Previously, people would have spent hours, months in doing all of this. Right? Now the promise with Nitro is that we radically make you efficient here.

Not incrementally, but radically.

There could be things like escalations, which are important signals that you have with existing customers.

If things are on track, you wanna know if your accounts are green or red. You wanna know if there's an expansion opportunity somewhere, or is some customer going to churn. You can have now real time insights into what's happening.

And then there's filling time sheets. No one has fun filling time sheets. So things like connect to my calendar, connect to my email, connect to my maybe GitHub, if you're a developer, and then figure out all the work that I did last week, and then fill that time sheet automatically in Rocketlane this week. It's just possible in ten minutes now.

You should be able to come into a system and just ask it. Hey. What are my revenue for the last three months? Are we on a upward or a downward trend? If it is downward, why is the trend downward? What went wrong?

One of our customers has implementation discovery questionnaire. It has three hundred questions. Can you imagine how much time it would take someone to fill this out? Now you point the calls to the document, and the agent fills it for you. You put in rules on how you want a question to be answered, and these are the only options which have to be picked and all of those things. Right?

So Imagine you just won a new deal and all the happy bells are going around in office. But there's one team which is now dreading what comes next, which is the implementation team because they know how much work these large onboardings can be. So what happens now for them is they will get into conversations with the customer. There's gonna be a lot of back and forth around what do they want, what exactly was sold, and a bunch of things around that. Right?

And then once you're through that, getting data from the customer, migrating it, setting up their instance, and so much is Then you sort of break down every part of the job and say, hey.

Why does that inefficiency have to exist? Can we make it faster or ideally zero?

For one of our customers, they used to spend around fifteen, sixteen hours on creating this thirty five page design document for each of their larger customers, and now that's become one and a half hours of work. Pulling content from multiple workshops they've been on with the customer and then emails that have been exchanged with the customer to create that document automatically. We have this account intelligence piece, which I think is very unique for services teams.

If you go into a particular customer's, you know, profile and there are people listed there, you go into one of them, you can actually even see what are some potential icebreakers for a conversation with them. So, like, what are their likes that we have discovered from calls they've been on, etcetera. So I think it was a nice touch to say, okay. You're gonna meet with Rahul. Right. Here, it is what Rahul likes and has told us in the past that they care about, and that helps me break into a better conversation.

Does that mean that Nitro is gonna come for a lot of people's jobs in PS?

I don't think the idea with Nitro is that it's gonna come for your jobs.

Are you saying that because you wanna be nice?

I don't wanna be canceled on the Internet. Yes. Okay.

I think people derive meaning from work, right, whether we like it or not. Maybe they have, like, a sweet spot on how much they would like to work. Okay. But I don't think the answer is zero.

Everyone wants to go somewhere, but I think the smaller teams would spend less time getting there. But at the same time, we are positioned really well because we are not too small, we are not too large. For us to change direction is quite easy, but at the same time taking into considerations things like SOC two, SOC one, and all the, you know we have ISO forty two thousand one now, which is an AI specific certification. So we have to keep those controls in mind while we make those changes. At the same time, we are not so large that it takes us a really long time to adapt to change in AI. Because in AI, if you miss a change over a week, it's like you lost a month.

So is Nitro the future of professional services? Maybe, maybe not. Time will tell. Having spoken to the people who actually built this out, one thing seems pretty clear to me. For years, software has been keeping track of work, and now software actually wants to do the work, which is exciting and also mildly concerning. On that note, this is Sam Toddler for Rocketlane News.

‍