Nextnet Copilot

Nextnet Copilot is an AI research assistant built specifically for life sciences teams. It delivers evidence-backed answers drawn from Nextnet's semantic web of biomedical data, which unifies scientific literature, gene and protein data, pharmacological insights, and more. Copilot is designed for researchers across biotech startups, academia, pharma, and enterprise settings who need reliable, hallucination-free answers without sifting through irrelevant content.

Copilot functions similarly to general-purpose AI assistants such as ChatGPT, Perplexity, or Microsoft Copilot, but restricts its sources to verified scientific works including research papers, literature reviews, reputable news, and press releases. Every response is grounded in real data from Nextnet's knowledge graph, with sources and specific excerpts cited alongside each answer to support confident, traceable research.

Core Capabilities

• Provides evidence-backed answers sourced exclusively from verified scientific literature and biomedical databases.
• Cites primary sources and pinpoints specific excerpts — the most relevant snippets — from each source.
• Suggests follow-up questions to support iterative, hypothesis-driven research workflows.
• Allows users to curate and filter sources by number of excerpts, citation counts, impact factors, and other criteria.
• Enables navigation from chat responses directly to all associated sources within the interface.
• Supports viewing full text of source documents via a "View Full Text" option.
• Reduces the risk of AI hallucinations through built-in guardrails and retrieval from a controlled scientific knowledge base.

Collaboration and Sharing

• Sessions can be shared with team members or external collaborators as guests.
• Export options include PDF download or shareable session links, distributable via email, Slack, or other channels.
• Organisation management tools allow administrators to control access and manage individual team spaces.
• Curated sources from a Copilot session can be handed off to Nextnet Explorer for deeper brainstorming and connected science mapping.
• Upcoming collaboration features include version control, change logs, and advanced privacy and access controls.

Underlying Technology

• Built on a custom retrieval-augmented generation (RAG) pipeline with multiple large language models (LLMs) chained together.
• Semantic web foundation uses knowledge graphs and vector databases to connect fragmented biomedical data.
• Incorporates up-to-date natural language processing (NLP) to derive semantic concepts and relationships across scientific literature.
• Uses a purpose-built ontology developed by scientists to create a common vocabulary and surface deeper conceptual relationships.
• Data is kept current through automated crawlers and ingestion pipelines.
• Runs on scalable cloud infrastructure.

Interface and Usability

• Designed with a clean reading interface for straightforward navigation between chat messages and sources.
• Sources icon in the upper right corner of the interface provides access to a full list of primary sources and excerpts from any session.
• Compatible with Google Chrome, Microsoft Edge, Safari, and Firefox.

Nextnet Copilot is available via Free and Starter subscription plans, with additional tiers for teams and enterprise users. It operates as a web-based platform and is used by researchers globally to accelerate literature review, hypothesis exploration, and R&D workflows.