SBTi + AI: Driving the Future of Corporate Decarbonization

Introduction

Welcome. Thanks so much for being with us today at this virtual webinar. We're so happy you've joined us, and I'm really excited to talk about this topic: SBTi, AI, and decarbonization, and share a bit about how Pulsora supports companies along this decarbonization journey.

Today we're going to cover a few different topics. We'll share a bit about what we're seeing in the market today with regards to decarbonization and where organizations are. We'll talk about SBTi 2.0, the draft that was released earlier this year, and how various AI use cases and technology can help support that journey. We will also share some really exciting examples of how the Pulsora platform leverages AI along those use cases to support and supercharge the effort that sustainability teams are facing with decarbonization. Finally, we will leave some time for Q&A. Feel free to pop any questions in the chat throughout today's conversation, and we'll have some folks monitoring the chat throughout.

My name is Jessica Mathis, and I am a product manager here at Pulsora. I've been working in the sustainability and technology space for the last eight years, and have spent a lot of time in the AI space, both on classical AI like machine learning and now a lot of focus on generative AI.

About Pulsora

For those who Pulsora may not be familiar to, we are an AI-first, comprehensive platform that was really built for use cases all across the sustainability value chain. So all things when you think of reporting, compliance, managing your data collection processes, your carbon management and decarbonization and performance metrics, all in one place in this unified and purpose-built platform. Specifically, I lead our carbon product.

The state of corporate decarbonization

Before we dive into SBTi 2.0, I want to hear a little bit from you. We're going to launch a quick poll, and we really want to hear about the state of corporate decarbonization in your organization today, and specifically how you're thinking about science-based targets and net zero targets and goals.

Some of the reflections and thoughts of what we've been seeing at Pulsora: thousands of organizations across the globe, many of yours included, have submitted net zero targets and goals. That is really ambitious and really exciting. That challenge grows as you progress in terms of actually reaching those goals. The driver to submit those goals, and the reason why so many companies around the world are increasingly sharing these types of goals even amidst scrutiny for these public goals, is really threefold pressure from key areas: pressure from stakeholders like across supply chains. Many of your customers may have new supplier sustainability programs that are increasingly required. Secondly, a lot of interest in competitor net zero goals and that aspect of sustainability. And there's a lot of consumer pressure today, which I'm sure many of you are thinking about and feeling, and maybe even yourself are contributing to in a really positive way, which is finding value in green products and services.

What is SBTi 2.0?

You may have heard a bit about SBTi in the news this year. SBTi, Science-Based Targets initiative, has been around for a long time, but this year they released a new draft version of the Corporate Net Zero Standard. It's basically the standard by which companies around the world think of how they should be setting those targets and goals, both from a near-term and long-term emissions goals perspective.

There are a few different phases to this 2.0 version. This is part of a standard operating procedure, so it's normal for these things to be updated over time. The current version is the 1.2 version. Earlier this year, the draft for 2.0 was released. There was a big public consultation phase. Many of your companies may have contributed to that, whether through working groups or through public survey.

After that draft was released and the consultation period closed, we're now in an intermediary period, about to start a Phase Two pilot. That Phase Two pilot is going to take the initial draft and leverage it with real data, real companies, to see how it might work in process. Because really, the whole purpose is to make this Net Zero standard valuable for you, useful, and efficient.

Sometime in 2026, the new 2.0 standard will be released. It may look very close to what the draft is, or it may look somewhat different. For any of the folks who answered that they have not yet set targets, if you're interested in setting targets with SBTi, the guidance today, because we have this expectation that the 2.0 draft will be finalized in 2026 and in effect in 2027, the recommendation is to just focus on those near-term 2030 targets. The 1.2 version that's currently in effect is going to be phased out by 2030. So just focus on those near-term targets, and for long term you can start to think about planning. Near term has an impact on your long term as well, but maybe wait until the final 2.0 to get those long-term ones in place.

The four-step SBTi journey

Throughout this journey, many of you who have already gone through the SBTi process to some extent might be familiar with this. There are four main steps along the journey.

The first is declare an intent. To share with SBTi that you are interested in setting a goal or target for net zero. Specifically with 2.0, there's a new categorization of A or B. It's essentially based on your industry and your company size. Whether you're a small business or a very large business in a high-emissions industry, you're going to have a different guideline to follow. There are also some specific industry standards SBTi has put out for sectors like financials and energy.

After the intent phase, you have about a year to submit targets. This is that process where I'm saying congrats for the folks who have validated targets, because this is a really arduous process that many companies today are still using spreadsheets to support. Part of that process is baselining your entire corporate emissions inventory, using primary data wherever possible, and now with 2.0 getting some assurance on that baseline. Then coming up with those targets, both near-term and long-term: how much you want to decarbonize, figuring out that ambitious yet achievable target. That is a really difficult process, and a lot of data goes into that modeling.

Specifically, something new about the 2.0 draft is that previously we had a joint Scope 1 and 2 target. Now they're separated, so you have a separate Scope 1 and a separate Scope 2, both near and long-term targets. For Scope 2, you can also have a separate target for your market-based emissions. A lot of details there under "submit target". Highly recommend going and reading the draft standard or some of the summaries available online to get a bit more in the weeds.

Finally, what I think is perhaps the most exciting thing about the 2.0 draft is reporting on progress. There's a new concept of beyond value chain mitigation. Beyond things that are in your immediate value chain, maybe in your industry, there's the opportunity to help support decarbonizing a mine or some other operation that may not necessarily contribute to your immediate value chain but supports industry as a whole. That's something now there's the opportunity through SBTi to get extra recognition for.

The other aspect is a climate transition plan. In this new world, if the draft is validated as it says, you'll have to put together a transition plan, which means to really take the time to be thoughtful about how you're going to reach those goals, not just declaring the goals but actually making a plan. That's perhaps the trickiest part and one of the biggest opportunities for technology and AI to play a role.

What Pulsora's carbon management does

In the Pulsora platform, carbon management has two main areas of focus or workflows. One is just establishing your inventory: coming up with your baseline, making sure that you're using and incorporating as much primary data as possible, automatically mapping to the most relevant emissions factors based on industry and location. The steps collect, measure, and audit are robustly managed on an enterprise-grade platform. There's a ton of efficiencies and risk mitigation that comes with that versus using spreadsheets, which is where most companies start. That makes sense. You want to make sure you have a sense of all of the things that go into it, and then moving to a technology platform can really add additional value and propel you to reach those goals.

The second aspect: okay, you have an inventory, you have your emissions, so what are you going to do with it? That's where we provide enabling workflows on Pulsora to help support reporting, sharing out with your stakeholders, communicating those intents and strategic initiatives. Providing analytics dashboards, performance insights to find your hotspots and make sure that those climate transition plans, for example, are prioritized on areas that are going to give you the most impact. And providing scenario analysis so you can play with various short- and long-term plans and portfolios of projects or pathways within those plans, and see how that can contribute to your overall goals.

We support customers across carbon and the rest of the sustainability realm. Within carbon management, these are the two areas that we play, and a lot of customers are using our software to support reporting on CDP, PCAF, and others, in addition to science-based targets.

AI use cases for decarbonization

AI is the really exciting hot topic of the day. What I'd like to see is how your sustainability teams are using AI today. There are a lot of different AI tools and technologies out there. Maybe some are ones that your entire company is using and your sustainability team has unique use cases on, or maybe some are specific to your team.

One other thing while I give a minute for folks to respond to the poll: I'd be remiss not to mention there's a lot of discussion of adverse impacts of AI on climate. Responsible and mindful use of AI is something that's really important to us at Pulsora, and something that's really important to consider as your teams are incorporating AI into your workflows. How to do that in a way that really adds value, more than anything. There are real concerns about data centers' impacts on energy, water, waste, emissions, and even just the communities in which they operate. At Pulsora, when we think about AI, we're trying to be extremely mindful of where it can really add value, and not adding AI unnecessarily just because people are talking about it.

Within AI, all the talk today is about generative AI and LLMs, but there are also other classical AI use cases that can play a role in supporting your decarbonization journey. We want to consider both in five key areas.

The first is anomaly detection. This is a traditional classical machine learning approach. You're looking for anomalies in a data set. This can be really useful around decarbonization and emissions management, especially as you're working towards limited assurance processes and uncertainty reports, and to support data collection as a whole.

The second is scenario analysis. Using AI to generate different scenarios and manipulate different variables on those scenarios, doing predictive analytics of what could happen in various situations and how to best mitigate. Really helpful for near- and long-term planning as part of your SBTi goals.

We also have agentic data collection. You may have heard a lot about agents. Beyond the ChatGPT prompt-and-response experience of interaction with LLMs or GPTs, with agentic data collection you can leverage an AI agent to think about a task and then go and perform that task and make some assumptions along the way, or collaborate with other agents. It's getting closer to that more helper-function or assistant type of AI use case.

Then we have forecasting, another classical machine learning approach. Really important when you think about decarbonization. Finally, data generation. Some folks mentioned in the poll that they're using AI for brainstorming and playing around with ideas. That's a great use of data generation. Another can be crafting a response to a particular question. Sustainability teams get inundated with different questionnaires, both internal and external, and AI can be really helpful at speeding that process up and making sure it's leveraging data from the right sources.

Demo: Anomaly detection

Here we are in the Pulsora platform. This is an example of what a task might look like. In Pulsora you can send a distributed data request across your entire value chain. It could be to within your organizational hierarchy like your facilities, or to suppliers, or to downstream customers and partners or distributors. In this example, the user has provided several different electricity utility invoices. They may have uploaded them in bulk or one by one, but we're able to provide really immediate anomaly detection insights.

You can see on the bottom that we have a couple of these bills that have the warning label against them. It looks like there's a really high spike in amount of energy for this month of September. For the user who may not understand the context around why this is shown as an anomaly, we also have this helpful visual to provide additional context and make sure that the user can see really easily, "Oh wow, September is a magnitude of 200% higher than these other months, so clearly there's some variance happening here." They're able to identify that anomaly at the point of data entry or data collection before this task is submitted for an approval workflow or mapped to emission factors and gets all the way downstream to contributing to your overall Scope 2 emissions number. With anomaly detection, we're going back to the source and making sure we're identifying any issues upfront, supporting the user with fixing them in the moment to prevent any downstream erroneous data points.

Demo: Decarbonization pathways

The second use case is our decarbonization pathways module on Pulsora. Within carbon management, we have this module that's helpful for both setting targets and taking action on them. We want to make sure we're able to operationalize those decarbonization goals given all of the data unified and collected under the platform.

Here you can see a list of projects that have been generated using AI to support the specific goals this organization has set out. You can see a couple, like solar panel installation or an advanced HVAC system. The AI is able to generate relevant projects, as well as generate the potential reduction amount, which is critical for a net zero manager who may be managing both a budget and a set of projects and trying to determine which projects to fund, which may have risk associated with them, which have a longer duration until they'll realize impact. This is the main place where all of this data can be managed. We're using AI to augment. You may already have some low-hanging fruit, some projects already in flight, but it's a great use case of AI to also generate additional project ideas to make sure you can action on your net zero goals.

Demo: Agentic supplier data collection

The next AI use case is related to data collection, agentic data collection. This is a supplier emissions dashboard. Something I want to point out specifically is our data quality score related to our supplier emissions details. Embarking on a supplier engagement program, some folks may even set SBTi goals around percent of suppliers that have set SBTi goals or are engaged on sustainability initiatives. You may want to reach out to those material suppliers, material meaning you spend a lot of money with them and they are contributing significantly to your GHG emissions.

Some suppliers are inundated with requests or may not have the technical capability or domain expertise to support a full emissions inventory. In lieu of requiring that, for any suppliers that are not responsive or don't know how to do this, AI can be really helpful at filling in the gaps. For example, if Supplier Six is not responsive in terms of their emissions, but there is another report that is available on the web and has stated that Supplier Six last year had a certain amount of emissions, then that agent can go web-search, find that publicly available information, and populate it into Pulsora.

We're keeping track of that here because it has a really high data quality score. Over time, you can see where the agent has intervened and where you've actually collected primary data, like the one score here with Supplier Five. Over time you can keep track of that and try to improve that level of engagement with those suppliers, and perhaps decarbonize alongside them. This is a really powerful use case.

Demo: Forecasting with Prophet

The next AI use case is around forecasting. We want to make sure that not only are you setting goals and creating projects to support those goals, but you're actually able to model the way those goals are going to change over time based on your business-as-usual or business forecasts, and based on seasonality. Within Pulsora, this is a sample analytics dashboard where we have both an emissions and water forecast for a global company. We're able to use different models. In this case, we're using the Prophet model by Meta, which incorporates seasonality into these impacts on emissions and water usage over time.

Demo: Report narrative generation

Finally, I want to talk about the use case for data generation. This may be the one most familiar if you're using ChatGPT and asking it "Hey, write me an email draft about this," or "Correct this paragraph to give me a one-line summary." Similarly, embedded in Pulsora without having to leave the platform, if someone is assigned a metric to provide, maybe your colleague is responsible for filling out a narrative portion of your impact statement or of an emissions report or a submission to SBTi, you can use Pulsora AI to actually draft that narrative response in kind of a Mad Libs format. That's going to make it much easier for that person to start from a draft, much easier than a blank sheet of paper, and be able to fulfill that request quickly and with relevant data. The model is also able to provide the contextual information they need if this is the first time that user has been assigned this data point.

What sets Pulsora apart

We've looked at five examples of AI use cases embedded within Pulsora workflows. Before Q&A, one thing I want to wrap up is what sets Pulsora apart. Three main takeaways to think about when you're pursuing those SBTi plans and looking for a platform or AI solution to help move from spreadsheets and emails to a more robust, future-ready approach.

The first is centralized data management. We talked about SBTi 2.0 and how no one really knows exactly what that final draft is going to look like. There are also several other frameworks you may be reporting against, like CDP or PCAF. To be able to collect and manage your data in one central place and flexibly apply those metrics across any number of frameworks, without having to duplicate that effort, is hugely powerful. Pulsora is designed to support that. All that data has full data governance, you can trust it, and you can flexibly apply it across different frameworks.

Second: transparency and auditability. You saw the example earlier about how we're applying AI in this space. Most of our customers are going through internal and external audit now. Pulsora has been built from the beginning with full end-to-end traceability. Nothing is a black box. You always know how the data has been touched, who's touched it, how and when and why. All of those things are really easy to find directly in the platform. Many audit firms actually log on specifically to view the data in a read-only format directly in their customer's environment. It makes the whole process very efficient.

Finally, the AI-powered platform. We have a lot more coming when it comes to AI use cases, and we're excited to continue to release them in very specific and value-added ways to supercharge your journey.

Q&A: Will validated SBTi targets need revision for 2.0?

If you have targets that are for 2030, you will not need to revise them for 2.0. If they're beyond 2030, like long-term 2050 targets or 2045 targets, you may have to revise them. Stay super close on those SBTi updates. That's something that Pulsora will also continue to provide updates on. We're going to continue to monitor, and based on the second phase pilot results at the end of this year and what we see released next year, we'll continue to provide those updates.

Q&A: Utility data integration options

There are several different ways to integrate data into Pulsora, and we work with many different systems or data lakes. We have pre-built integrations with several different sources, which could include things like Utility API. You can also include bulk file uploads into your process. If there's a way to get it out of that utility data provider, or if you have an aggregator already, you can upload those files directly. We also have invoice scanning, which we're really excited about launching very soon. That's something AI-powered as well. You can also use SFTP-type integration. If there's a file folder you want to populate, we can securely do that via secure file transfer.

Q&A: Permissions for value chain respondents

If you're using Pulsora to reach out directly to outside external folks across their value chain, those folks, if they're responding to a data request, can access the platform totally free. It is something we can work with you to set up in terms of how we get those email addresses or how we set up their account. It's fairly seamless, and we can make sure that they are integrated into your supplier engagement journey. We'll actually be coming out with some more specific supplier engagement pieces later this year, so keep an eye out for those.

Q&A: Data security

Pulsora has quite a few certifications: SOC 1, SOC 2, and GDPR compliance. Several of our customers have very high data security requirements that we've had to pass, so we're very confident in the level of data security within our platform. In terms of any of the AI use cases, we are specifically leveraging models that have a no-training policy. Any data that our customers are sharing through our Copilot or any of the features on the platform is not going to leak or be shared publicly in any way, or be used for any LLM training.

Q&A: High-growth companies with low baselines

Even within SBTi, there is no expectation that you'll have 100% absolute reduction. There is some understanding now that carbon credits and offsets have a role to play in this. So that's totally accepted and part of the plan. If you're a young high-growth company and you are smaller, you're probably going to fall under that A and B category, which has a specific designation for you. It would actually be specific to both your company size and your industry. SBTi has specific standards that are going to apply to your company profile. They may not be exactly the same as a really large company that may have a much larger global footprint.

In terms of your point about growth, that's definitely important. That's why those forecasting features and other aspects on Pulsora can really help you make sure you're not setting out targets that are not achievable or not aligned with your business growth. Something I've also seen really popular and valuable, a good place to start, is an intensity-based metric. Thinking about your emissions intensity per revenue dollar and trying to reduce that over time can sometimes be a good entry point.

Q&A: Leveraging CDP and GRI data for SBTi

The answer is yes. Anything that is relevant from those frameworks that you can leverage for SBTi, you don't have to recollect or remanage that data. A lot of those emissions baseline metrics, and things like that, and goals are going to be easily portable over. Some of the other more specific SBTi requirements you may want to augment to be able to submit for validation. The great thing is, if you're already reporting to CDP and GRI, you are most of the way there. You have a really strong foundation and good starting point, and you can continue to just focus on the gap and then submit to SBTi.

Q&A: CBAM

CBAM is an important regulation that is coming. CBAM is super important, especially to very specific industries and the emissions associated with products in those industries. Pulsora does not have a product carbon footprint module or an LCA module. That's something we would collaborate with another provider on. But we would be very much part of that reporting, that EPD process for CBAM, and helping to manage your corporate carbon inventory aspect of it.

Closing

That is all we have time for. Thank you guys so much for joining today. I hope this was valuable. We will be sharing some follow-up information via email and the link to the webinar in case you want to send it to a colleague or someone who wasn't able to join. We really value your time and your questions, and we'll follow up with the remaining questions offline. Thank you so much.

‍