In late 2020, I founded Nextnet to organize and integrate the world’s scientific knowledge and make it accessible. We wanted to enable researchers to rapidly generate insights and unlock limitless possibilities. 

Over the past four years, we have relentlessly pursued this mission by building, testing, designing, analyzing, learning from failures, and refining our approach based on insights from thousands of users and potential customers.

In early 2024, we launched our Minimum Viable Product (MVP). The response was immediate. Through organic word-of-mouth, our user base grew by an astonishing 1,200% within a year. 

Today, Nextnet is used in over 100 countries. While half of our users are based in North America and Europe, many others come from places such as Brazil, India, and Israel.

Researchers spend hours navigating fragmented biomedical knowledge sources. Whether searching through PubMed, Google Scholar, bioRxiv, patents, grant reports, or policy documents, they find themselves buried under irrelevant information and disconnected datasets. 

I’ve experienced this firsthand in my career, first as a student in New Zealand, then as a visiting scholar at UC Berkeley, and then while founding my first biotech, Mekonos.

Scientific inquiry requires seamless evidence and prior knowledge integration. Yet, existing systems fail to provide a cohesive solution despite researchers’ best efforts. Examples such as the Unified Medical Language System (UMLS) and Medical Subject Headings (MeSH) have limited scope and low interoperability. The Open Biomedical and Biological Ontology (OBO) Foundry, which encompasses over 500 ontologies with millions of entities, struggles with duplication, inconsistencies, and poor connectivity, making large-scale research applications cumbersome and inefficient.

Beyond semantics, life sciences are plagued by disjointed data repositories. Open science initiatives have generated countless data sources but lack standardization and cross-compatibility. 

Many repositories partially reference or duplicate others, but without robust mechanisms for seamless cross-querying, researchers must piece together fragmented datasets manually. And applying AI to such incomplete data sets can lead to biased and unreliable results, limiting the transformative potential of AI in life sciences.

While software has eaten much of the world, life sciences and healthcare lag far behind. 

Most available software in the life sciences industry comes from two places. The first are usually expensive, rigid point-solution SaaS applications built by outsourced IT consulting firms. Secondly, ad-hoc tools developed by graduate students using research grants. These tools often lack proper documentation, are outdated, or serve as single-purpose solutions.

For researchers, knowledge and workflows remain trapped in disparate silos. At best, they resort to duct-taping solutions using email, Slack, MS Office 365, SharePoint, and pen & paper. Their lack of access to comprehensive knowledge management leads to costly human errors and uninformed decision-making.

Another challenge shared by researchers is team collaboration. Research roles are now more cross-functional than ever. They must juggle many tasks like running experiments, reviewing scientific papers, analyzing clinical trial results, sharing insights, writing reports, and strategizing commercial efforts. 

While software engineers have numerous tools to work together as one team, researchers still struggle with collaborative workflows. No single tool streamlines their workflow effectively for research, discovery, reading, commenting, sharing, and storing.

Scientists and researchers crave high-quality tools to rely on for their mission-critical work. After 18 months of close collaboration with our MVP users and beta customers, I’m confident we’re well positioned to meet their needs.

Think of Nextnet as an AI knowledge companion built to help life sciences teams discover, visualize, and share research. 

While an exceptional user experience is key, our real breakthrough lies in coherently linking information about a subject matter across disparate data contexts at scale and then visualizing the connections. 

For example, when a user queries Nextnet about therapies, the platform doesn’t simply return isolated search results. Instead, it contextualizes knowledge from scientific literature and other fragmented sources. The result is a persistent institutional memory of research. This continuous connection of ideas enables reanalysis and discovery beyond conventional search tools.

Our purpose-built ontology is at the core of Nextnet’s infrastructure. It is designed to make scientific data intuitive and actionable. 

Instead of working with raw tables and rows, Nextnet understands real-world biological entities and the causal relationships between them. It provides an interactive environment where users can engage with data intuitively, rather than relying on complex queries or code.

Users don’t need to know SQL or employ engineers to write code to search petabytes of data. Instead, Nextnet does the heavy lifting:

• Nextnet Copilot: Delivers precise, evidence-backed answers to mission-critical questions in real time. So you can focus on breakthrough discoveries.
• Nextnet Explorer: Dynamic visual workspaces to help researchers map complex ideas, uncover hidden connections, and navigate research intuitively.

Beyond search and visualization, I’m even more excited about unlocking new possibilities for team collaboration. Whether you’re sharing insights instantly via a link, giving contextual feedback on research findings, or setting shared workspace access to foster seamless teamwork, Nextnet makes it easy to work together. Check out more in the demo below.

Research teams can now gain early access to the new platform features. Access is free, making it easy for individuals and small teams to start using Nextnet immediately. 

In the next few months, we will introduce Premium plans in addition to the Free plan. 

Coming up next, Free and Premium plans will be supported by valuable features such as:

• Organizations & team spaces: Enhanced access and privacy controls for organizations using Nextnet. 
• Robust sharing controls: Easily share any resource in your Nextnet organization with guests and teammates.
• Guest access: Offer external access to your Nextnet organization and team spaces.
• Favorites: Bookmark your favorite Copilot chats and Explorer sessions. 
• Collaborative comments: Conveniently located directly within an Explorer map.
• More data sources: Expanding your research with data from clinical trials, grants, and patents. 
• Global search: Quickly find resources stored in your Nextnet organization. 
• Disappearing mode: Copilot and Explorer without data retention on the Nextnet platform. 
• More intuitive Explorer: Not sure where to start? We’re building more intuitive features in the Explorer to help you navigate the most relevant data entities. 
• Quotas: A generous portion of usage limits included in the free plan.
• Open in Explorer: Deep-dive integration between Copilot and Explorer.
• Search filters: More robust control of results.
• Results expansions: Iterate through a larger set of search results in ranked order.
• Expand data nodes: Go deeper into the data by revealing connections not initially shown.
• Document upload: Expand your research by bringing your research papers and other documents into Copilot. 
• Mobile: Introduce core features within mobile browsers and iOS/Android apps.
• Speaking mode: Interact with Copilot using audio. 
• Reader: A more robust reading experience of research articles. Navigate by excerpts and semantic concepts. 
• Pages: To iterate content across discoveries in Explorer sessions, Copilot chats, external information, and created content. Pages is the space for more complete knowledge management and a foundation for your organization’s institutional memory.

Let me know what you think about the features roadmap! Your opinion will help us prioritize.

The possibilities with Nextnet are endless, and we can’t wait to see what you’ll discover. 🚀