The most expensive mistake in construction isn't a wrong measurement or a delayed shipment. It's the collective, optimistic bias baked into every project schedule. For nearly a decade, London-based nPlan has been quietly building a machine to counteract that bias, not with more rules, but with data. Its core asset is a proprietary dataset of over 750,000 completed construction programs, representing more than $2.5 trillion in capital spend [nPlan, October 2025]. The company's software analyzes this history to forecast where and when new projects are most likely to stall, turning retrospective post-mortems into forward-looking risk assessments.
Founded in 2017 by Dev Amratia and Alan Mosca, nPlan operates in the narrow but critical wedge of schedule assurance. Its primary customers are owner-operators and major contractors executing large capital projects in sectors like energy, water, and transportation. These are organizations for whom a multi-week delay can cascade into millions in cost overruns. nPlan's pitch is that its AI-led forecasts provide an objective, data-driven counterweight to the inherently optimistic planning of human teams [nPlan, “Our Story”].
The data moat and the product wedge
nPlan's defensibility starts with the dataset. Aggregating and normalizing hundreds of thousands of construction schedules,each with its own taxonomy, software format, and regional peculiarities,is a significant data engineering challenge. The company claims this corpus allows its models to identify patterns of delay that are invisible at the single-project level. For instance, it can surface that electrical rough-in tasks for a specific type of industrial facility have historically overrun by 23% on average, or that certain sequences of trades are statistically prone to congestion.
The product suite is built around this predictive engine. nPlan Insights is the core web-based software for single project teams, offering risk forecasts and opportunity identification [nPlan, Products Overview]. nPlan Portfolio extends this to program managers, integrating with tools like Power BI to provide a portfolio-wide risk dashboard [nPlan, Products Overview]. A newer product, Schedule Studio, aims further upstream, using generative AI to help planners convert a scope document into a logically linked construction schedule in under an hour, according to the company [nPlan, Blog Posts].
Traction with regulated giants
nPlan's growth has been methodical, targeting the slow-moving but deep-pocketed end of the construction industry. As of 2021, the company operated in eight countries with nearly 30 customers across commercial, infrastructure, transportation, and energy construction [For Construction Pros, March 2021]. Recent named customers point to a focus on regulated utilities and major infrastructure owners. Chevron Technology Ventures became an investor in the 2025 Series B, and the oil major is also a customer [The Standard, June 2026]. Other disclosed clients include Anglian Water, Sizewell C, and National Grid [The Standard, June 2026].
This customer profile aligns with the product's value proposition. For a utility like Anglian Water, which is embarking on a multi-billion pound investment period (AMP8), the ability to de-risk a portfolio of hundreds of concurrent projects is a material financial safeguard. nPlan serves these clients by providing what it calls "investment case assurance" and unbiased schedule assessments, often used to evaluate competing contractor bids [nPlan, Solutions Overview].
Funding a global deployment push
nPlan has raised a total of approximately $34.5 million in disclosed funding to build and deploy its platform. A significant $18.5 million Series A in 2021 was led by GV, the venture arm of Alphabet, signaling early confidence in the technical approach [For Construction Pros, March 2021]. The most recent capital was a $16 million Series B in October 2025, led by CapHorn with continued participation from GV and new strategic investors Chevron Technology Ventures and Suffolk Technologies [nPlan, October 2025].
The funding history shows a progression from proving the core AI to scaling commercial deployment, particularly in North America.
2021 Series A | 18.5 | M USD
2025 Series B | 16 | M USD
The Series B proceeds are earmarked for scaling the company's global operations, with a visible hiring push in the US market for deployment and sales roles [University of Toronto Entrepreneurship Job Board, October 2026].
The technical breakdown: how it works
From an infrastructure perspective, nPlan's system is a classic example of applied machine learning where the model is only as good as the pipeline feeding it. The technical workflow likely involves several distinct phases. First, ingestion and normalization of schedule data from various sources (Primavera P6, Microsoft Project files). Second, feature extraction to translate schedule logic, task durations, and resource assignments into a model-ready format. Third, the training of ensemble models on historical outcomes to predict probabilistic delays for new schedule inputs. Finally, the serving layer that delivers those forecasts through a web UI or API integration.
The most immediate risk at scale isn't model accuracy, but data pipeline integrity. As nPlan onboards more clients, the variance in schedule quality and the volume of real-time updates will increase. Maintaining a clean, normalized dataset against a torrent of noisy, client-generated input is a continuous engineering challenge. A single mis-mapped task dependency or incorrectly parsed calendar could propagate errors. Furthermore, the model's predictions are only actionable if project teams trust them, which requires not just statistical performance but also explainability,a persistent hurdle for complex ML systems.
What to watch in the next twelve months
The next phase for nPlan will test its transition from a novel risk-analysis tool to an embedded system of record. Key milestones will be landing a first nine-figure enterprise contract and proving renewal rates at that tier. The company's partnership with Chevron is a promising beachhead in the energy sector, but replicating that success with other Fortune 500 owner-operators will require demonstrating a clear return on investment, likely measured in delay prevention and contingency cost reduction.
Competitively, the landscape is fragmented but could consolidate. While no direct, named AI competitors are cited in sources, nPlan competes with incumbent project-management software suites and large consultancy practices that offer schedule risk analysis as a service. Its defensibility hinges on the unique scale of its historical dataset and the automation of insights that were previously manual and expert-dependent. The bet is that this automation creates a step-change in cost and speed, moving risk review from a quarterly audit to a continuous process.
The sober assessment is that nPlan's technical approach is sound, but its commercial success depends on navigating the conservative procurement cycles of its target industries. The software can identify where a trade is statistically likely to stall [Slashdot, 2026], but convincing a project director to reallocate resources based on that alert requires a cultural shift. The company's recent funding and strategic investor base suggest it has the capital and the partnerships to try. The next year will show if the industry is ready to trade gut feeling for a data-driven forecast.
Sources
- [nPlan, October 2025] nPlan raises $16M Series B to scale its AI-led transformation of capital project delivery | https://www.nplan.io/press-releases/nplan-raises-16m-series-b-to-scale-its-ai-led-transformation-of-capital-project-delivery
- [nPlan, “Our Story”] Our Story | https://www.nplan.io/our-story
- [For Construction Pros, March 2021] Construction Planning Software Startup nPlan Secures Funding From Google Ventures | https://www.forconstructionpros.com/construction-technology/news/21342976/nplan-construction-planning-software-startup-nplan-secures-funding-from-google-ventures
- [nPlan, Products Overview] Products Overview | nPlan | https://www.nplan.io/products-overview
- [nPlan, Blog Posts] Blog Posts | https://www.nplan.io/blog-posts
- [The Standard, June 2026] Inside nPlan: the AI fixing construction’s $500 billion delay problem | https://www.standard.co.uk/business/smes/inside-nplan-the-ai-fixing-construction-s-500-billion-delay-problem-b1285599.html
- [nPlan, Solutions Overview] Solutions Overview | https://www.nplan.io/solutions-overview
- [University of Toronto Entrepreneurship Job Board, October 2026] Deployment Engineer, AI Project Controls (UK) | https://jobs.entrepreneurs.utoronto.ca/companies/nplan/jobs/97399295-deployment-engineer-ai-project-controls-uk
- [Slashdot, 2026] nPlan software can identify where and how long trades, spaces or tasks have stalled | https://slashdot.org