Charlie User Guide

The AI Scientific Research Agent. Learn how to use Charlie effectively to accelerate your research, analyze scientific literature and maximize your discoveries.

Emerit Science Team

Updated: January 2026

Interface Charlie - Agent IA de recherche scientifique

Charlie is a scientific research agent powered by a large language model (LLM) and vector search across scientific article abstracts, primarily from PubMed.

It helps you accelerate your research, analyze scientific literature and explore new avenues of inquiry through artificial intelligence. This guide walks you through getting the most out of Charlie step by step.

1. How Does the Search Work?

When a question is asked to Charlie, here is a summary of what happens behind the scenes:

1 It takes into account the conversation history and adjusts its search if needed
2 It compares your question against PubMed article abstracts
3 It selects the 20 most relevant abstracts
4 It generates a response based on the content of these references
5 It displays the references and links to the sources used to answer

2. How to Interact with Charlie?

Charlie is designed to converse naturally, like a thinking partner. It understands both open-ended and specific questions, and you can structure your research approach step by step.

lightbulb The richness comes from dialogue

The more you exchange with Charlie, the more relevant and tailored the responses will be to your context

Steps for a Good Interaction

calendar_today
Use the year filter
Before asking your question for better temporal focus
chat_bubble_outline
Ask an initial question
Even if it's broad, Charlie will help you refine it
gps_fixed
Let Charlie suggest
Leads and angles of exploration you might not have considered
search
Analyze the response
And the sources provided to validate the information
sync
Dig deeper or pivot
Explore a specific point or discard certain leads

3. Best Practices

To optimize your use of Charlie and get the best results, here are some essential tips:

gps_fixed

Focus on one objective

One objective per session for better consistency in your research and more relevant results

calendar_today

Use temporal filters

Restrict sources to a more recent period to get up-to-date information

palette

Organize with colors

Use colors in the interface to categorize your thematic sessions

menu_book

Follow your scientific watch

Add relevant references to your workspace to easily come back to them

4. Example Questions

Here's how to progressively refine your thinking by exchanging with Charlie. These examples illustrate different interaction strategies.

Progressively deepening a subject

→ "What are the known roles of the TGF-β protein in breast cancer?"

→ "And specifically in the regulation of the tumor microenvironment?"

→ "Are there inhibitors targeting this pathway currently in development?"

Redirecting a discussion

→ "I was interested in the role of p53 in apoptosis, but could you instead summarize its involvement in cellular aging?"

→ "Actually, could you compare the functions of p53 with those of FOXO3?"

Specifying a research objective

→ "What strategies are used to deliver siRNA into tumor cells?"

→ "And more specifically in the case of brain tumors?"

→ "Are there approaches using nanoparticles?"

Exploring multiple angles

→ "What are the reported effects of metformin in oncology?"

→ "Are these effects observed independently of diabetes?"

→ "Can you list the mechanistic hypotheses proposed in the literature?"

Ready to accelerate your scientific research?

Start using Charlie for free and discover how AI can transform the way you conduct research.

Start for Free