Intel / article

Research Assistant: AstraZeneca's Agentic System for R&D

Source: arxiv.org
Author: Piotr Grabowski, Mohamed Alameen, Jorge Bretones, Sabina Cardell, Miguel Carmona, Gavin Edwards, Ben Grainger, Sameh Hassan, Erik Jansson, Artur Kuziakhmetov, Albert Maristany, Hebatallah Mohamed, Andriy Nikolov, Sebastian Nilsson, Mark O'Donoghue, James Pacileo, Ashiq Sultan, Alex Voegele, Michael Ughetto
Date: Published Aug 15, 2026

Context

AstraZeneca's Research Assistant brings biomedical evidence into a chat interface for the pharmaceutical company's scientists and clinicians. Its technical-note abstract describes retrieving information from sources including scientific literature, clinical trials and internal experimental systems, then linking answers back to the underlying material.

The system offers direct answering and a multi-step mode for more complex research questions. The useful design choice is that users can follow an answer back to its evidence and investigate further, rather than receiving an unsupported summary. The abstract describes deployment in internal research workflows but provides no quantified improvement in research productivity or clinical outcomes.

Terms in this piece · Glossary

LLM — A large language model — the neural network behind tools like Claude and ChatGPT, trained on huge amounts of text to predict what comes next.
grounding — Tying a model's answers to checkable sources — retrieved documents, live data, tool results — instead of letting it answer from memory alone.

Key quotes

“We describe Research Assistant, an internal LLM-based system developed at AstraZeneca to help scientists and clinicians explore biomedical questions across a broad range of data sources.”
Piotr Grabowski et al.

“The system provides a chat-style interface that brings together evidence from scientific literature, knowledge graphs, chemistry, clinical trials, safety resources, expression data, and internal experimental systems.”
Piotr Grabowski et al.

“It supports both a fast mode for direct question answering and a multi-step mode for more complex research tasks.”
Piotr Grabowski et al.

“Responses are grounded in retrieved evidence and linked back to the original sources, allowing users to review and further explore the underlying data.”
Piotr Grabowski et al.

Read the source arxiv.org