Watershed has launched a new set of AI agents designed to automate some of the most time-consuming tasks in sustainability work, while also introducing a new AI Fellowship program aimed at helping sustainability professionals build stronger internal AI capability. The company is positioning the launch around a familiar bottleneck in ESG and climate teams: too much time is still spent collecting, cleaning, reconciling, and analyzing data before any strategic decision-making can begin.

The significance of the announcement lies in how directly it targets that problem. Rather than presenting AI as a broad assistant layer, Watershed is focusing first on two concrete workflow areas: data cleaning and data analysis. That makes the launch more practical than conceptual, especially for sustainability teams still overwhelmed by fragmented source data and year-end reporting cycles. This is an inference based on the product focus described in the announcement.

 

The first product push is aimed at cleaning messy sustainability data faster

 

Watershed says its new agents can transform messy operational data into measurement-ready information in a single session. The tools are designed to handle unit conversions, date formats, country codes, duplicate records, missing values, broken formulas, and overlapping datasets, while also documenting assumptions and preserving review controls. According to Watershed, test customers saw time to actionable data fall by 80%, and one company reduced a five-hour data cleaning task to 20 minutes.

That matters because data preparation remains one of the least visible but most resource-intensive parts of sustainability work. In many organizations, climate and ESG teams still spend more time fixing data than using it. If the time reduction claims hold at broader scale, the value of the product may lie less in reporting speed alone and more in freeing teams to work on decarbonisation, procurement, and strategy. This is an inference based on the use cases described by Watershed customers.

 

Read more: Siemens and Vulcan Energy Partner to Scale Europe’s Sustainable Lithium Supply With 24,000-Ton Lionheart Project

 

The analysis agents aim to turn emissions data into strategic insight

 

Alongside cleaning, Watershed has introduced data analysis agents that allow users to ask natural language questions about emissions data, methodology, and underlying reporting logic. The company says the tools can deliver year-over-year comparisons, identify decarbonisation hotspots, flag anomalies, and provide drill-down explanations tied back to source data. One early customer said the analysis agents surfaced an insight that otherwise would have taken three analysts and multiple weeks to generate.

This is important because sustainability software is increasingly expected to do more than store data and produce disclosures. The next competitive layer is likely to be whether platforms can turn structured and unstructured data into decision support. Watershed is clearly trying to move in that direction by linking AI to strategic analysis rather than only workflow automation. This is an inference based on the product positioning and example customer outcomes.

 

Customer examples show the product is being sold on time recovery and decision speed

 

Watershed’s launch relies heavily on real customer workflow examples. Royal Mail’s climate strategy team reportedly used the agents to process refrigerant and travel data in minutes instead of hours, while Smiths Group said the tools are helping save about 12 weeks per year and reducing monthly Scope 1 and 2 reporting work from about a week to a day or less.

These examples matter because they show how sustainability AI is increasingly being framed in business terms. The sales case is no longer only that AI helps write reports faster. It is that time recovered from reporting and data preparation can be redirected into emissions reduction, energy efficiency, and strategy execution. That shift makes the commercial logic of sustainability AI easier for companies to justify internally. This is an inference based on the customer quotes included in the launch.

 

Watershed is also trying to solve the sustainability AI skills gap

 

The company is pairing the product launch with the Watershed AI Fellowship, an eight-week program running from May 12 to June 30, 2026. The fellowship is open to Watershed customers and is designed to help sustainability professionals work directly with Watershed’s AI product leaders, test workflows, shape new features, and build responsible use cases for sustainability AI inside their organizations.

That addition is strategically important because Watershed cites a survey of more than 200 global sustainability leaders showing that 37% see internal skills gaps as a limiting factor in adopting AI for sustainability. In other words, the company is not only selling software. It is also trying to create a user base with enough capability to embed AI into sustainability operations more deeply. This is an inference based on the survey result and the structure of the fellowship.

 

Explore OneStop ESG Marketplace: ESG Software

 

Watershed is positioning itself as a purpose-built sustainability AI platform

 

A core part of the company’s argument is that general-purpose AI is not enough for sustainability work because the field requires auditability, regulatory awareness, methodological rigor, and data lineage. Watershed says its agents are built on sustainability-specific databases containing more than 500,000 emissions factors covering 95% of global GDP, and that outputs include changelogs, lineage, and hallucination checks.

This matters because trust and defensibility are likely to be the main dividing line between general AI tools and enterprise sustainability AI platforms. In reporting and climate data management, speed alone is not sufficient if outputs cannot be reviewed, defended, or audited. Watershed is clearly using that requirement as a differentiator. This is an inference based on the company’s product architecture claims.

 

The broader signal is that sustainability AI is moving into a more mature phase

 

The launch suggests sustainability AI is entering a more operational and specialized stage. Companies no longer seem to want AI only for drafting narrative reports or summarizing disclosures. They want it to reduce manual data burden, improve analytical speed, and help teams act more strategically with the information they already have.

For Watershed, this announcement strengthens its attempt to lead that category by combining workflow automation, analytics, audit-ready infrastructure, and user capability building in one offer. For the wider market, it signals that the next phase of sustainability software competition may be shaped less by who has AI features and more by who can make sustainability data genuinely usable at speed without sacrificing rigor. This is an inference based on the positioning of the product and fellowship together.

 

Source: Watershed