Q1/2025 Product Updates: Laying the Groundwork for Open Science: OpenDeSci’s Product Progress

BLOG POST

Berlin, Germany – 13th April 2025 

As we build momentum at OpenDeSci, we’re reflecting on the foundational work we’ve completed and the strategic direction ahead. Our mission remains clear: to reshape how scientific knowledge is discovered, organized, and shared—using decentralized technologies, AI, and scalable infrastructure.

---

Beta Testing with selected users (register for waitlist): app.opendesci.org

---

This update outlines what we’ve built at OpenDeSci, the insights we’ve gathered through direct engagement with the research community, and the systems we’re developing to scale discovery and impact in the DeSci space.

Strategic Focus: Research, Validation, and Prototyping

We began this quarter by validating our hypothesis: that the infrastructure supporting today’s scientific research ecosystem is fragmented, slow-moving, and inaccessible to many researchers, particularly in emerging domains or interdisciplinary fields.

To explore this, we focused on three core initiatives:

1. Early Prototyping to Explore Product-Market Fit

We developed and launched an interactive prototype that allows users to explore emerging research topics, surface relevant content, and begin to understand how AI-powered knowledge graphs can support discovery. This prototype is a first step in our journey to create a semantically-rich, continuously updating platform that reflects the evolving landscape of global science.

2. In-Depth User Research with Scientists and Domain Experts

Over the course of the quarter, we conducted interviews with researchers across domains including genomics, neurobiology, bioinformatics, and computational medicine. These conversations revealed recurring pain points such as:

• Fragmented access to scientific information across journals, repositories, and informal networks

• Difficulty staying updated on new papers, especially in interdisciplinary fields

• Limited discoverability of early-stage or unpublished work (e.g. preprints, protocols, datasets)

• Minimal visibility into the people behind the research—their motivations, collaborations, and current focus

These insights have helped us prioritize the capabilities we’re building—focusing not only on papers, but also on the people, trends, and questions driving the research ecosystem forward.

3. Intelligent Crawlers as Proof of Concept

We implemented intelligent web crawlers to serve as a proof of concept for continuous content discovery. These crawlers are capable of identifying new scientific outputs (papers, news, protocol updates, and domain-specific resources) and categorizing them in real-time using agentic AI pipelines. This is the backbone for our future recommendation and alert systems—enabling a new layer of machine-assisted discovery.

---

Infrastructure and Product Stack: From Prototype to Scalable Platform

By the end of Q1, we transitioned from a simple prototype to a fully cloud-hosted and scalable application infrastructure. Our technology stack, tools, and infrastructure choices are designed for performance, global availability, resilience, and developer collaboration.

Figure 1: OpenDeSci Cloud Architecture

• Hosting: All services are deployed on AWS, providing global scalability, reliability, and compliance with security standards.

• Backend: We use FastAPI, a modern Python web framework optimized for asynchronous operations, enabling high-performance APIs and efficient request handling.

• Frontend: Built with Vite and React, our frontend ensures fast load times, efficient development, and a responsive user experience.

• CI/CD: Deployments are fully automated through continuous integration and delivery pipelines, enabling rapid iteration and stability.

• Architecture: The system is modular and horizontally scalable, supporting growing workloads across both research and commercial use cases.

This setup supports rapid deployment, scalable testing, and sustained performance as our user base expands. Further, on this platform, we’re now live with our first set of user-facing features. These are designed to support exploration, research discovery, and scientific networking.

1. Topics Explorer

The Topics Explorer allows users to browse and explore research domains across multiple fields such as genetics, neuroscience, AI, longevity, biotech, and more. Each topic surfaces new and relevant papers, researchers, and subfields through real-time content feeds and AI-assisted tagging. This is our first step toward a semantic map of global science.

Figure 2:  OpenDeSci Topics Explorer Feature

2. Researcher Profiles

We’ve introduced dynamic researcher profiles to provide a richer understanding of who is driving scientific progress. Profiles include associated publications, institutional affiliations, collaboration networks, and research themes. This supports not only discovery, but also the development of community and trust within the OpenDeSci ecosystem.

Figure 3:  OpenDeSci Researcher Profiles Feature

3. AI-Powered Science Shorts

We’ve launched an early version of AI-generated research summaries—short, digestible overviews of scientific topics, papers, and developments. These summaries are designed to make complex topics accessible while highlighting the key contributions of a given paper or research trend.

Figure 4:  OpenDeSci AI-Powered Science Shorts Feature

Looking Ahead: Toward an Open Graph of Scientific Knowledge

As we enter Q2, our focus will shift toward building the first version of our semantic knowledge graph. This open infrastructure connects researchers, research outputs, and open questions using graph-based AI models. This is the foundational architecture for enabling dynamic search, personalized alerts, and intelligent recommendations.

We will also expand our AI crawler system to include additional data sources such as preprint servers, protocol repositories, grant databases, and GitHub repositories for scientific code.

Ultimately, our vision is to enable a transparent, intelligent, and collaborative research ecosystem—where the discovery of knowledge is accelerated by open data, open tools, and open communities.

Beta Testing with selected users (register for waitlist): app.opendesci.org

If you are a researcher, innovator, or builder in the space, we would love to connect and learn from your perspective. Reach out directly or follow our work as we continue to scale.

—

Neil Desh

(CTO, Product & Engineering)

Neil is a seasoned tech innovator with extensive experience in full-stack development, cloud infrastructure, and AI-driven solutions. His experience in the academic space began when developing QiWord, a research platform leveraging NLP to organize and search academic publications. More recently, he played a key role in the success of StellaConnect, contributing to its $100m+ acquisition by Medallia. His dedication to driving innovation through cutting-edge technology makes him an invaluable leader in the tech industry.