New product announcement

Introducing nPlan AutoReport - project controls reporting made easy

Today we're launching nPlan AutoReport, a new standalone tool which uses a world-first Generative AI system specifically designed for the needs of project controls reporting. Consider this your essential guide to our newest game-changing product.

Written by
Carlos Ledezma
Carlos is a Product Manager at nPlan. His work focuses primarily on the use of Generative AI functionality to super-charge the user experience of nPlan users. He also has extensive background in Machine Learning, which he uses to inform the technical aspect of the features that he works with.

The goal with everything we build here at nPlan is two-fold: firstly we want to give project professionals valuable insights they can use to make smart decisions, and secondly we want to help our customers turn their forecasting and risk management process into a powerful lever they can pull on to increase their chances of delivering on time and on budget. 

Today we are delighted to announce the launch of nPlan AutoReport, a Generative AI-powered tool which delivers on both of these ‘value propositions’ for our customers. Users of the tool will gain the ability to meet their reporting commitments in a fraction of the time previously required, freeing them up to spend more time proactively managing risk, and thereby streamlining and improving the efficacy of their forecasting and risk management process. 

In addition, by utilising the latest developments in Generative AI, users will be able to extract valuable insights from their data more easily, and rapidly generate reports communicating those insights to the right stakeholders, enabling more of the smart decision-making alluded to above. 

But how did we get here? Where did our journey to AutoReporting start? Let’s dive in…

Regular project reporting: time-consuming, boring - and important

A major capital project is a huge enterprise with hundreds--if not thousands–-of employees, and the involvement of many different interacting companies, each with their own realms of responsibility and processes. Reporting artefacts serve as a means to keep everyone up to date about what is going on and escalate any issues to the relevant stakeholders so that appropriate action can be taken. Moreover, regular reporting serves as an audit trail; when projects fall behind, for example, there is evidence that shows how incidents were managed, which risks were considered at the planning stage, and which part of the project has driven the (most) delay. However, while project reporting undoubtedly serves an important function, it is also incredibly time-consuming and unloved. 

We regularly interview project controls professionals to gain a better understanding of where and how we can best help them with their work. While doing this type of research, one common pain point relates to reporting on changes to a risk register - we often hear how this process is time consuming and “not what I became a risk manager for”. Upon digging deeper, we discovered just how long the list of reporting artefacts that different people in a project controls team produce on a regular basis actually is - and the amount of time spent on reporting activities is, frankly, staggering.

We found that it can take up to two weeks of somebody’s time to produce a monthly risk report for a major capital project. That is 50% of someone’s working time spent solely on reporting duties! People are also not thrilled about having to do so much reporting. During our research, we were repeatedly told by interviewees that their “job was supposed to be to manage risks, not just report on them”. Now, two weeks per month is an extreme case that one encounters in large projects with heavy reporting requirements. Even so, our research found that project managers, planners and risk managers spend, on average, anywhere between 10% and 15% of their time on reporting-related tasks. These tasks could be gathering data from different teams, evaluating the data sources, or writing up the reports in slide decks or documents.

Don’t get me wrong, there are very efficient teams in the project controls world. One project manager told us that they could put a monthly progress report together in under one hour with their team, provided that they had all the data available. However, throughout our research it became clear that in the average case, reporting is a time-consuming task.

Perhaps more importantly, we have yet to meet a project controls professional who loves sitting down for a few hours to do some good ol’ progress reporting. So as we said above, it is clear that while regular reporting serves an important function, it is incredibly time-consuming and often gruelling. What could we do about it?

The elements of a useful report: an opportunity for Generative AI 

What makes a good report? Based on our user research, the answer is as follows: 

  1. A focus on insights derived from a range of data sources. Having a spreadsheet with a formula that calculates a metric is easy. It’s harder to take data from different departments (scheduling, cost, risk, etc.) and weave that data into an accurate narrative about the progress of the project.
  2. Tailoring for specific audiences The industry has defaulted to “catch-all” reports that are sent out to all stakeholders; then the stakeholders need to decide for themselves what in the report is important for them to understand. However, we found that people - especially more senior people - prefer shorter reports that contain just the data that is relevant to them.
  3. Replicability. All progress reports have a regular cadence, with a template that needs to be filled in. These templates are typically created towards the beginning of the project and change rarely. So the main job to be done is to gather and interpret the data, using the template as a guideline.

At face value, creating reports that meet these criteria is a job well suited for Generative AI. The models behind Generative AI are quite capable of taking various data sources, summarising them and extracting insights from all the information that they have been provided with. 

However, anyone who has interacted with GenAI before will know that it’s just not that simple. There are hallucinations to handle, prompt engineering - and the biggest question of all: how do you ‘give’ project controls information (e.g. schedules, risk registers, health and safety reports, etc.) to the Generative AI solutions that are commercially available nowadays so that they can work their magic? 

nPlan is uniquely placed to solve these challenges. We already have internal methods we use to read schedule data and risk registers in a way that enables AI models to understand them. On top of that, we have a web platform that supports a user experience tailored to project controls. If we built a Generative AI-powered reporting tool, our users wouldn’t have to take all the data required for reporting to a tool that is optimised for chatting, they could simply use the data they already have inside our tool which is specifically designed for project controls - and create reports in a frictionless way.

We could see that we had a golden opportunity to make a positive difference to the working lives of project professionals, and began work on a new Generative AI-powered reporting tool based on the following principles:

  1. User experience above all: we have all seen too many chatbots that are hard to prompt, and that hallucinate. Our tool would guide the user through the process of creating a report and use the latest techniques to avoid generative hallucinations.
  2. Specific to project controls: our tool would work out of the box for the project controls community and be optimised for the queries and file types common to project controls.
  3. Collaborative and iterative by default: reporting is a collaborative endeavour, and we didn’t want to change that. Collaboration, personalisation and ease of recreating reports with new data would be elements of our tool.

With our plans well and truly laid, we set to work. And the result, nPlan AutoReport, is now ready for project professionals to use. In the next part of this article, we’ll lay out the features of AutoReport and demonstrate exactly how it’ll transform your reporting life. Let’s go!

nPlan AutoReport, Powered by Barry

We’ve built a product that will reduce the time it takes to meet regular reporting obligations from days to minutes. The data that goes into a report can be uploaded directly to the tool, and it’s stored there for re-use in any new report that you need. The tool offers flexibility to generate and edit content in different AI-assisted ways. Collaboration is the default mode, with anyone in the team able to log into the tool and contribute to any report. Whenever a report is ready, you can export it in one click. And perhaps most valuable of all, AutoReport allows you to template the reports you build, and reuse them whenever a report needs to be recreated with new data. 

Crucially, you don’t need to know anything about Generative AI to make the most of AutoReport!

Build reports from your project files - whatever format they’re in.

We have designed our AI agents to understand the data formats that are most common in project controls. Users can simply drag and drop files and don’t have to worry about whether the AI will understand them. On release our AI agents will be able to understand:

  1. Schedule files exported from the major planning tools: Primavera P6, MS Project and Asta PowerProject.
  2. Risk registers exported from any risk register manager in spreadsheet format (CSV, XLS, XLSX)
  3. Any text-heavy document that forms part of your regular project progress in PDF format. Examples of these documents include early warning notices, health and safety reports, meeting minutes and risk reports from contractors.

These will be available without needing any processing or data curation. Just drag and drop via our interface, and they will be ready for use in reports.

Our AI agents are capable of understanding most documents in project controls

Build reports one section at a time; generate text, tables and charts with ease

We have built our user experience in a way that allows the user to build one section of the report at a time. This means that users have maximum flexibility to build reports in any format that they need, whilst at the same time minimising the amount of prompt engineering that they have to do. 

Each section is created using a guided experience where the user can select which documents specifically should be used for that section and they can use two different types of Generative AI: one will allow them to generate text-based content (e.g. summaries, comparisons, narratives, tables, etc.), and another which will allow them to generate charts from descriptions in natural language.

By using generative AI to support writing one section at a time, users get maximum flexibility to create their reports whilst minimising the amount of prompt engineering they need to do.

Collaborate to create

The focus on individual sections also serves to highlight collaborative report writing. If a report consists of multiple sections that require multiple sources of expertise, each member of a team can log into AutoReport, upload their data and contribute as many sections as required. Changes made by any user are visible to all, as soon as they’re made.

Turn reports into templates, and generate reports in minutes

One of the most powerful features of our reporting functionality is that once all the sections of a report are created, the report is automatically available to be used as a template. So the following month, when the user needs to create the report again, all they have to do is upload the new data, duplicate the report, and wait a few minutes for their report to be ready.

Given that we are using Generative AI, we encourage users to review and make any adjustments or corrections required to the content that has been generated. This is very simple to do with our rich text editor, which enables the user to edit sections or delete them altogether.

Overall, we have aimed for an experience that makes it seem like our users are not interacting with generative AI models. Instead, our technology has been designed to make the user feel as if they’re simply following a step-by-step process to create complex reports in collaboration with the top minds in their company.

How to get your hands on AutoReport

AutoReport is available as a standalone product and as a module within Insights Pro. Users of both versions will have access to the easy reporting functionality and document reading capabilities of the tool; the main difference is that users of the standalone product will not have access to the forecasting and risk analysis data that is available in Insights Pro. As those familiar with nPlan will know, this forecasting and risk analysis data is generated by Predictive AI (specifically Deep Learning models) trained on over 750,000 schedules from past projects - and as such it is incredibly valuable for guiding capital project decision-making.

Watch this space for a future blog on the power of combining AutoReport with Insights Pro to produce reports that look into the future of a project - as well as summarise its past.

Get AutoReport for your project today

AutoReport increases productivity, reduces boring work and helps project controls professionals focus on the creative management tasks that have the biggest impact on project outcomes. So if you’re interested in spending a lot less time reporting, and a lot more time doing, contact us at to find out about bringing AutoReport, Powered by Barry™, to your project today. Want to know more about the technology? Check out our new AutoReport product page.