Introducing Devano AI

We started Devano with a simple but ambitious mission: use modern AI to solve the everyday challenges that impact the pace of scientific discovery.

In the life sciences, data exists in a fascinating ecosystem of interconnected knowledge. Biomedical information naturally spans diverse formats, sources, and systems—reflecting the rich complexity of biology itself. This creates a fundamental challenge: questions that should be straightforward to answer often require hours of tedious manual work.

A researcher looking to compile "all human kidney studies using single-cell RNA-seq" faces a deceptively complex task. This information might exist across Gene Expression Omnibus (GEO), PubMed, supplementary materials, and other sources, but connecting the dots means manually:

• Sifting through free-text fields that might describe "kidney," "renal tissue," or "nephric samples"
• Cross-referencing papers to find crucial sample metadata not included in repository entries
• Determining experimental methods from inconsistently formatted protocol descriptions
• Understanding the access patterns and limitations of each data source

This manual connection-making steals time from the creative, high-value work that advances science.

Throughout our careers, we’ve experienced this pain firsthand. We have also continually seen how traditional automation fails to solve the connectivity problem because it lacks the flexibility to bridge across inconsistent terminology, recognize implied relationships, and adapt to the evolving landscape of biological research.

Today, we have a powerful new class of tools—large language models and AI agents—that can navigate this complexity, recognize patterns across modalities, and learn from domain expertise and feedback.

At Devano, we believe this changes everything.

We're building the intelligence platform for modern biotech—a system that empowers researchers to extract, curate and enhance complex biomedical data with speed, precision, and trust. By combining powerful agentic AI workflows with thoughtful design, we're helping researchers spend less time navigating data structures and more time advancing scientific understanding.

Our First Product: AI-Augmented GEO Curation

Our first product is built around the Gene Expression Omnibus (GEO), a foundational repository for gene expression data. While GEO is rich in content, it's also notoriously complex. Designed as a general-purpose archive, its utility often depends on how individual researchers choose to format and submit their data. As a result, metadata is frequently inconsistent across studies, and critical sample-level information is often buried in supplementary materials or hidden in spreadsheets. For the many scientists who rely on GEO, this turns the process of identifying and characterizing studies into a time-consuming detective mission.

We built Devano’s GEO platform to change that. With our GEO and PubMed agents, users can:

• Run AI-enhanced searches that go beyond keyword matching by dynamically extracting the most relevant latent metadata
• Automatically extract, structure, and harmonize metadata and sample-level information
• Track evidence and maintain a clear audit trail for reproducibility
• Interact with results through a clean, purpose-built web UI

In collaboration with a forward-thinking design partner, we’ve deployed this system in production—and it's already on track to save their teams hundreds of hours of manual curation work.

What once took days now takes minutes.

Why “Devano”?

We named our company Devano AI after the Spanish word devanar. Its primary meaning is "to wind," like the coiling strands of DNA 🧬 that we've spent so much of our careers working with, and which Devano AI's life-science customers often work with. But devanar also has an idiomatic meaning: "to rack one's brain," as in Yo me devano el cerebro ("I rack my brain"). We love the image of tireless AI agents racking their brains on complex tasks, freeing you up for more creative and enjoyable work.

What kind of company do we want to be?

In a space as complex and competitive as biotech, how we work matters just as much as what we build. Our values guide everything we do — from the way we design and ship products to how we collaborate with scientists, partners, and each other. They help us stay grounded, aligned, and focused on what truly matters. So we’re putting them front and center—as a reminder to ourselves and a signal to anyone who works with us.

Science is Our Compass

We believe that science is humanity’s best method for understanding reality and improving life. We build tools that support and accelerate the scientific process, and we hold ourselves to the same standards of rigor, transparency, and continuous learning.

Solve Real problems

We’re not here for status or buzzwords. We focus on meaningful problems that move science forward and improve people’s lives.

Data Should Be a Bridge, Not a Barrier

Organizing biological data is really hard, and that leads to challenges to researchers in finding exactly what you're looking for. We’re committed to building systems that unify, clarify, and unlock scientific data—making it more usable, actionable, and trustworthy.

AI with Purpose

We view AI not as a gimmick, but as a powerful augmentation of human insight. Our work is grounded in domain expertise, humility, and a belief that intelligent systems should make scientists faster, sharper, and freer—not replace them.

Flow Is Key

Great tools disappear. We believe in building products that are joyful, elegant, and seamless—experiences that help scientists enter a state of focus, clarity, and creative output.

Craft Matters

Thoughtful design, clean code, and technical excellence aren’t just aesthetics—they’re how we earn trust, reduce friction, and enable better science.

Truth Over Ego

We care more about getting it right than being right. We lead with evidence, challenge assumptions (especially our own), and prioritize collaboration over competition.

Be Kind, Be Real

We operate with honesty, humility, and respect—toward our teammates, users, and partners. We believe kindness and candor are not in conflict. They’re how we build trust and momentum.

What's Next?

We’re just getting started. We’re continuing to expand our agent architecture, deepen integrations with biomedical resources, and partner with more teams to unlock the next generation of AI-powered data workflows.

If you’re a scientist, developer, or biotech leader who’s tired of spending hours on manual glue work—we’d love to connect.

Let’s build a better way forward for science.