Skip to content

Prompt

I'd like to use LLMs to suggest ways in which I could perform more effectively at work. Something I've been looking into is creating a repository of files whose purpose is providing key chunks of context for LLMs. Can you suggest how I might go about building up a context repository for this specific purpose?

Output

Building a context repository to enhance LLM performance for work-related tasks is an excellent approach. Here are some suggestions for creating and organizing such a repository:

Content Types to Include

  1. Work Processes and Procedures: Document your company's standard operating procedures, workflows, and best practices.
  2. Project Documentation: Include summaries of past and current projects, lessons learned, and successful strategies.
  3. Performance Metrics: Compile key performance indicators (KPIs) and metrics used in your role and organization.
  4. Industry Knowledge: Add relevant industry reports, trends, and analysis specific to your field.
  5. Company Policies: Include important company policies, especially those related to performance and productivity.
  6. Skills and Competencies: Document key skills and competencies required for your role and industry.
  7. Tools and Technologies: Create guides on the tools and technologies you use regularly.
  8. Meeting Notes and Action Items: Summarize key points from important meetings and track action items.
  9. Personal Development Plans: Include your goals, areas for improvement, and development strategies.
  10. Feedback and Reviews: Store performance reviews and feedback you've received.

Organization and Structure

  1. Categorization: Organize files into logical categories (e.g., "Projects", "Skills", "Policies").
  2. Tagging System: Implement a tagging system to allow for cross-referencing and easier retrieval.
  3. Version Control: Keep track of document versions, especially for frequently updated information.
  4. Summarization: Create concise summaries for longer documents to aid quick retrieval.
  5. Regular Updates: Set a schedule to review and update the repository regularly.

File Formats and Storage

  1. Text-Based Formats: Use plain text or markdown files for easy parsing by LLMs.
  2. Structured Data: For metrics or quantitative data, consider using CSV or JSON formats.
  3. Cloud Storage: Use a cloud storage solution for easy access and sharing if needed.
  4. Searchability: Ensure your storage solution allows for easy searching and indexing.

Implementation Tips

  1. Start Small: Begin with the most critical information and gradually expand.
  2. Consistency: Maintain a consistent format and structure across documents.
  3. Accessibility: Ensure the repository is easily accessible when you need to provide context to an LLM.
  4. Privacy and Security: Be mindful of sensitive information and ensure appropriate security measures.
  5. Integration: Consider how you'll integrate this repository with your LLM workflow (e.g., through a retrieval system).

By building and maintaining such a context repository, you'll be able to provide rich, relevant information to LLMs, enabling more accurate and tailored suggestions for improving your work performance. Remember to regularly review and update the repository to keep it current and valuable[1][3][6].

Extracted Prompts

Prompts/Creating a context for professional purposes.md#Prompt 1

The above text was generated by a large language model (LLM) and its accuracy has not been validated. This page is part of 'LLMs-on-LLMs,' a Github repository by Daniel Rosehill which explores how curious humans can use LLMs to better their understanding of LLMs and AI. However, the information should not be regarded as authoritative and given the fast pace of evolution in LLM technology will eventually become deprecated. This footer was added at 16-Nov-2024.