KnetMiner
We joined KnetMiner as their product team to update an old codebase and with time add AI to their research platform. Every AI feature can be turned on or off for a team without a new release, and it keeps working from its last saved state even if the AI service goes down. Now researchers can just ask the assistant to set up the graph view, pull together the right papers, or summarize a gene page. People who aren't scientists like growers can spot a plant disease from a single photo. And teams can build their own reusable search shortcuts without writing any code.
About project
KnetMiner is an integrated, intelligent, interactive gene and gene network discovery platform that helps scientists explore and understand the biological stories of complex traits and diseases across species.
Product is a research platform built out of Rothamsted Research that helps plant scientists find links between genes and traits across species like wheat, rice and Arabidopsis. The tool was already live and used by researchers every day, but its gene-network view had been built years earlier and was running on outdated tech. KnetMiner's small founding team needed a partner who could modernize this inherited codebase without ever taking the tool offline and stay on as a long-term extension of their team for frontend, backend and design. That's where Empat came in, and over time we also added a production AI layer on top of their knowledge graph.
KnetMiner is an integrated, intelligent, interactive gene and gene network discovery platform that helps scientists explore and understand the biological stories of complex traits and diseases across species.
Product is a research platform built out of Rothamsted Research that helps plant scientists find links between genes and traits across species like wheat, rice and Arabidopsis. The tool was already live and used by researchers every day, but its gene-network view had been built years earlier and was running on outdated tech. KnetMiner's small founding team needed a partner who could modernize this inherited codebase without ever taking the tool offline and stay on as a long-term extension of their team for frontend, backend and design. That's where Empat came in, and over time we also added a production AI layer on top of their knowledge graph.
Challenges we faced
Our first challenge was to modernize the codebase and rebuild the gene-network visualization engine without ever taking the tool offline.But the platform had a deeper problem. KnetMiner is powerful, yet hard to use: finding gene–trait links means navigating a dense filtering interface by hand and reading through scattered literature. Answers exist, but getting to them takes real effort and expertise - which puts the tool out of reach for non-specialists such as growers and agronomists entirely.
That's why AI was needed: to let scientists simply ask a question instead of manually configuring the graph, and to open the platform up to a wider audience. The harder half of the work, then, was making that AI dependable enough for daily research every feature safely controllable, with regulatory-grade data security, predictable costs, and graceful fallback if the AI service went quiet.
Solutions we delivered
Extend the capabilities of your existing full time employees and grow your business with a dedicated offshore team.
What about your case?
PROJECT IN DETAILS
KnetMiner's assistant is network-aware, not a generic chatbot: every answer is grounded in the specific gene network on screen and comes back with clickable citations to the exact nodes and edges it drew from. Its Agentic Search goes further, letting the model call tools directly to run multi-step gene discovery across connected sources. Underneath sits a full AI layer built end-to-end - authentication, encrypted history, usage quotas and streaming that turns a raw model into a governed, research-grade feature set safe to run in production.
More relevant cases
%201.avif)
.avif)
.png)
