Decluttr

Inspiration

Decluttr was inspired by the massive, mostly invisible footprint of dark data, with roughly 65% of all stored data never actually used, and estimates suggesting that if we cleaned up this wasted information globally, the resulting energy savings could power the worlds electricity needs for around fifty years. We saw a disconnect between what companies say about corporate social responsibility and sustainability and what they actually measure in their digital infrastructure, especially around data centers and storage that quietly burn energy for no real purpose. Dark data rarely shows up in ESG reports even though it drives emissions and costs without creating value, so we wanted to bring that hidden problem to the front of the conversation. Decluttr was born from the idea that deleting unused data should be treated like an easy high leverage climate action, giving companies a way to back up their sustainability claims with something concrete.

What it does

Decluttr is a data sustainability and intelligence platform that helps companies reduce their environmental impact by identifying and eliminating inefficient data usage across workplace tools. It aggregates data from platforms such as email, storage, and collaboration systems, and analyzes it to identify dark data, duplication, and unnecessary storage consumption. The platform computes a comprehensive data health score and uses AI to generate clear diagnostics and environmentally conscious recommendations. Through an interactive interface, employees gain visibility into their individual contribution to company-wide data waste, enabling them to take meaningful action. By making digital waste visible and actionable, Decluttr helps organizations reduce energy consumption, lower their cloud carbon footprint, and actively support their sustainability and CSR initiatives.

How we built it

We built Decluttr as a scalable, full-stack platform designed to model and optimize digital sustainability across organizations. The backend was developed using FastAPI, providing a modular and high-performance API layer for data processing and analytics. Data structures were defined using Pydantic to ensure consistency and extensibility. The frontend was built with React and Tailwind CSS, featuring an interactive graph-based visualization that represents different data sources as nodes within a unified ecosystem. We integrated OpenAI to power the intelligence layer, enabling the system to generate context-aware diagnostics and sustainability-focused recommendations. The architecture is designed to be modular, allowing seamless integration with enterprise tools such as Google Workspace, Slack, and cloud storage providers.

Challenges we ran into

One of our biggest challenges was figuring out how to translate the abstract idea of digital sustainability into something measurable inside an app, especially when we needed to estimate energy use and carbon emissions based on storage patterns without access to every detail of a companys infrastructure. We also had to design a flow where recommending data deletion feels safe and trustworthy, since people are understandably afraid of losing important files, so we needed thoughtful risk levels, confirmation steps, and clear context for every suggested action. Integrating IT concerns like security and compliance with sustainability goals in a single interface was tricky because those groups usually speak very different languages and care about different metrics. On top of that, we were working under hackathon time pressure, constantly balancing our ambition for features like more accurate emissions modeling with the need to ship a working prototype by the deadline.

Accomplishments that we're proud of

We are proud that Decluttr turns the hidden issue of dark data into something visual and actionable instead of just another line in a sustainability report, giving teams a concrete map of where their digital waste lives and what deleting it could accomplish. It feels really satisfying to see the stack come together, from the Python FastAPI backend and Pydantic models to the React and Tailwind dashboard and the OpenAI powered explanations, all working in sync instead of feeling like disconnected tech demos. For a short hackathon window, getting from a rough concept about dark data to a tool that can actually scan sample datasets, flag low value information, and show estimated impact feels like a big achievement. We are also proud that we managed to frame the project in terms of both cost savings and climate responsibility so it can appeal to engineering teams, finance, and CSR groups at the same time.

What we learned

We learned just how large the problem of dark data is and how much of the worlds storage is taken up by information that no one looks at but that still consumes energy and generates emissions in data centers. That changed the way we think about the cloud from something abstract and weightless to something with a real physical footprint and real climate consequences behind every extra terabyte. From a technical perspective, we gained experience in building APIs that need to be both precise for developers and understandable for non technical users, and in thinking about data governance and risk when your core feature is recommending deletions. We also learned a lot about how to frame sustainability features so they are not just nice to have add ons but are tightly connected to core business benefits like cost reduction and easier reporting for ESG targets.

What's next for Decluttr

Next, we want to connect Decluttr to real world storage systems like common cloud providers and document platforms so companies can onboard quickly and see dark data insights on their actual infrastructure instead of sample datasets. We plan to refine our impact models with better benchmarks so our energy and emissions estimates get closer to what sustainability teams need for formal reporting and climate goals. We also want to deepen the corporate social responsibility layer with richer dashboards, exportable ESG ready reports, and integrations into existing sustainability tooling so that digital decarbonization becomes part of their standard workflow. Longer term, we see Decluttr evolving from just a storage cleanup tool into a broader digital sustainability companion that helps organizations continuously monitor and reduce the footprint of their data and compute, not just perform a one time cleanup.