Only 10,000 out of 400 million companies worldwide are currently measuring their carbon footprints, according to Singapore-based climate technology start-up Unravel Carbon.
With climate change accelerating, Unravel Carbon offers an AI-powered carbon decarbonisation platform to help large and medium-sized companies track and reduce their carbon emissions. The platform converts a company’s accounting information into full supply chain carbon insights before suggesting climate solutions towards net-zero emissions.
Offering that capability, however, requires datasets to be unified. This is why Unravel Carbon leverages Snowflake’s Data Cloud to consolidate datasets so that customers can report on their carbon footprint within seconds or minutes. For comparison, traditional approaches of carbon reporting require six to nine months of manual consulting work.
"With Snowflake's simple-to-use platform, we can build and harness databases that identify emission hotspots quickly, while releasing less carbon than legacy solutions. We are also able to pull structured and unstructured data from various sources to drive effective decarbonisation initiatives,” says Grace Sai, co-founder and CEO of Unravel Carbon.
She continues: "In the three months since we integrated Snowflake's solutions, our emissions databases have been instrumental in improving data-driven, climate-aware decision-making. Unravel Carbon also is planning to set up the world’s largest repository for every service on Snowflake to power the low carbon economy.”
Sanjay Deshmukh, Snowflake’s senior regional vice president for Asean and India, adds: “We believe sustainability is very much a ‘Data Problem’. We are thrilled to collaborate with Unravel Carbon in this journey, leveraging data and data analytics and artificial intelligence (AI) for good to help more companies across the Asia Pacific become more sustainable.”
See also: Testing QA New Section BDC Feature Winner 1
Success factors for AI
Besides Unravel Carbon, many organisations are adopting some forms of artificial intelligence (AI) to enhance their operations or drive business growth. At a media briefing last Friday, Deshmukh shares the three things organisations must do to fully benefit from their AI deployment.
The first is to be able to unify all types of data. “It is important to understand that AI is powered by data. So the first step in getting more value from AI is to break down data silos and bring all your data together into one single source truth like onto the Snowflake Data Platform. [Only then] can you train models on top of that to leverage AI,” he says.
Secondly, organisations need to keep security at the front and centre of their AI and data strategy. “Traditionally, organisations that started their AI journey early tend to dump their data in the cloud. Then their data scientists would take those data, and put it on their machine to train and build AI models. This means sensitive customer and business data is in unsecured environments, and if that data is compromised, organisations will lose customer trust,” claims Deshmukh.
See also: Unpublished article shouldnt be accessible testing
He adds that Snowflake’s platform offers many capabilities that help organisations put security at the core of their AI strategy. “We allow tokenisation, [data] masking and multiple other strategies to help our customers keep the data secure [regardless of the user and applications].”
Finally, Deshmukh highlights the need to democratise AI. He explains: “We see different levels of competencies and data literacy within the customer organisation. On one hand, there are data scientists who are very comfortable using code and building AI models. Our job is to make sure that we let data scientists use the power of Python and other programming languages to build and train models.
“On the other extreme, there are citizen scientists, who are using auto machine learning models or pre-built models [on cloud platforms] for predictive analytics. We’ve [recently] enabled the Microsoft AI capabilities and machine learning capabilities on our platform. And we've been doing that with Amazon Web Services for many years. [By doing so,] we are truly democratising AI and machine learning by letting the user choose the tool of their choice.”
