> ## Documentation Index > Fetch the complete documentation index at: https://docs.raycaster.ai/llms.txt > Use this file to discover all available pages before exploring further. # Custom Variable Resolution > Harness the power of Raycaster's custom variable resolution using the '@' key to scale your research

Custom Variable Resolution

Leverage the '@' key to dynamically access and utilize data across your workspace for powerful, context-aware analysis.

@

Dynamic References

Reference any data point across your tables with simple '@' syntax

Contextual Intelligence

Build smarter prompts that adapt based on your data context

Workflow Automation

Streamline complex processes with variable-driven templates

## What is Custom Variable Resolution? Think of Custom Variable Resolution as a smart way to "fill in the blanks" in your documents and templates. It's like having a personal assistant who knows exactly where to find all your information and can instantly place it wherever you need it. ### In Simple Terms Custom Variable Resolution is a feature that lets you insert information from your data tables into your documents or messages by simply typing the '@' symbol followed by the name of the data you want to insert. **Real-World Example:** It's similar to how mail merge works in word processors, but much more powerful. Instead of just inserting names and addresses, you can pull in any piece of information from any table in your workspace.

Everyday Analogy

Imagine you're writing a birthday card to a friend. Instead of writing "Happy Birthday" generically, you could write "Happy \[AGE] Birthday, \[NAME]!" Now imagine if those brackets automatically filled with the correct information for each friend. That's what Custom Variable Resolution does, but for all your business data.

## Why Use Custom Variable Resolution? Here's how Custom Variable Resolution helps you in everyday tasks: 1. **Save time** - Stop copying and pasting data between documents and tables 2. **Reduce errors** - Automatically insert accurate information without typos 3. **Personalize at scale** - Create customized documents for hundreds of contacts with a single template 4. **Keep information updated** - When your source data changes, all references update automatically 5. **Create smarter templates** - Build documents that adapt based on the data they contain ## How Custom Variable Resolution Works - Step by Step
1

Type the '@' Symbol

Start by typing the '@' symbol wherever you want to insert data from your tables. This tells Raycaster you want to reference some information.

2

Choose What Data to Insert

After typing '@', a helpful menu appears showing you all the tables and data fields you can choose from. Simply select the information you want to insert.

3

Raycaster Fills in the Data

When your document is used, Raycaster automatically replaces all the '@' references with the actual information from your tables.

4

Your Document Changes with Your Data

As your data changes in your tables, any document using that data through '@' references will automatically update to show the latest information.

## How to Use the '@' Symbol - Simple Examples Here are some basic ways to use the '@' symbol to pull in your data:

What you type:

I'm writing to @ (search menu appears) Company about your services.

What gets inserted in your document:

I'm writing to {"{{"}Company{"}}"} about your services.

What the AI sees:

I'm writing to Acme Corporation about your services.

When you type "@", a search menu appears, allowing you to find and select the variable you need. Once selected, it's displayed as {"{{"}Company{"}}"} in your document, but will be replaced with the actual value when viewed.

Variable resolution interface showing picker, values, and debugging tools ## Real-World Examples - How People Use Custom Variable Resolution ### Example 1: Research-Focused Patent Analysis

Template:

Subject: Analysis of {"{{"}patent\_number{"}}"} patent

I need to analyze the {"{{"}patent\_number{"}}"} patent titled "{"{{"}patent\_title{"}}"}" filed by {"{{"}inventor{"}}"}.

This patent appears to cover innovations in the {"{{"}technology\_field{"}}"} domain, particularly regarding {"{{"}key\_innovation{"}}"}.

Please provide:

1. A summary of the main claims in this patent

2. An analysis of how this compares to similar patents in the {"{{"}technology\_field{"}}"} space

3. Potential applications beyond what's explicitly stated in the patent

Please focus your analysis on how this might impact research in {"{{"}research\_focus{"}}"}.

Filled Example:

Subject: Analysis of US10984584B2 patent

I need to analyze the US10984584B2 patent titled "Method and system for neural network based image enhancement" filed by Sarah Chen.

This patent appears to cover innovations in the computer vision domain, particularly regarding real-time image enhancement using adaptive neural networks.

Please provide:

1. A summary of the main claims in this patent

2. An analysis of how this compares to similar patents in the computer vision space

3. Potential applications beyond what's explicitly stated in the patent

Please focus your analysis on how this might impact research in medical imaging diagnostics.

### Example 2: Academic Publication Research

Template:

Research Analysis: {"{{"}publication\_title{"}}"}

Authors: {"{{"}authors{"}}"}

Journal: {"{{"}journal\_name{"}}"} ({"{{"}publication\_date{"}}"} )

DOI: {"{{"}doi{"}}"}

Keywords: {"{{"}keywords{"}}"}

Research Questions:

1. How does the methodology in this paper compare to {"{{"}comparison\_paper{"}}"}?

2. What are the key limitations identified in this study?

3. How might these findings apply to {"{{"}application\_field{"}}"}?

Synthesis Task:

Integrate these findings with our current understanding of {"{{"}research\_topic{"}}"} and propose potential next research directions.

Filled Example:

Research Analysis: Large Language Models as Zero-Shot Reasoners

Authors: Kojima, T., Gu, S.S., Reid, M., Matsuo, Y. and Iwasawa, Y.

Journal: Nature Machine Intelligence (2023)

DOI: 10.1038/s42256-022-00593-2

Keywords: large language models, reasoning, zero-shot learning, chain-of-thought

Research Questions:

1. How does the methodology in this paper compare to Brown et al. (2020) on few-shot prompting?

2. What are the key limitations identified in this study?

3. How might these findings apply to medical diagnosis assistance?

Synthesis Task:

Integrate these findings with our current understanding of prompt engineering techniques and propose potential next research directions.

## Getting Started Today Ready to try Custom Variable Resolution? Here's how to get started: 1. **Start simple** - Begin by creating a basic template with just one or two '@' references 2. **Experiment** - Try different types of references to see how they work 3. **Create templates** - Once you're comfortable, create templates for your common tasks 4. **Share with your team** - Help others save time by sharing your useful templates Custom Variable Resolution is one of the most powerful ways to save time and ensure consistency and scale your work with Raycaster. The '@' key is your gateway to smarter, more efficient documents.