How to leverage Robin’s document memory

What is Document Memory?

Document Memory transforms how you interact with your contract repository by bringing enhanced AI-powered search capabilities to Robin's Legal Intelligence Platform. This feature allows you to ask natural language questions about your contracts and receive comprehensive insights with specific citations to source documents. Unlike traditional label-based searching, Document Memory analyses the actual content of agreements, helping you synthesise provisions across fund documents, side letters, and portfolio company agreements.

Please note: Document Memory is currently in beta and only available to a limited group of customers. While powerful, all responses should be verified against source documents as we continue to improve this functionality.

‍

How to Access and Use Document Memory

1. Navigate to Query in the Robin Platform: Access Document Memory by logging into Robin and selecting the Query module from the top-left navigation bar.
2. Filter Your Document Set: Before you can ask questions, narrow your scope using Query's labelling capabilities:
   • Group: Select Group in the drop down and filter for a subset (e.g., "Fund III")
   • Filter by Contract Type: Choose categories like "Side Letters" or "Credit Agreements"
   • Select by Counterparty: Filter by specific investors or portfolio companies by leveraging Text Search:
3. Initiate Document Memory: After setting your filters, click the "Chat with Robin" button at the bottom right of your screen.
4. Ask Your Question: Type your question in the chat window. Document Memory will analyse all documents within your filtered selection.
5. Leverage Clickable Citations: Verify the accuracy of AI-generated answers by examining your documents and leveraging clickable citations.‍
6. Refresh for New Topics: When switching to a completely different topic or after changing your label filters, refresh the page or clear the chat to ensure accurate results that reflect your current selection.

‍

Effective Question Types for Private Equity

1. Analyse Terms Across Multiple Documents

Use Case: Quickly analyse complex terms across multiple agreements to provide timely answers about restrictions, obligations, and compliance requirements.

Example Prompt: ""Which of our credit agreements contain change of control provisions that would be triggered by our proposed restructuring, and what are the notification timelines for each affected lender?”

2. Managing Critical Event Response

Use Case: When significant events occur, rapidly identify all contractual obligations, notification requirements, and remediation timelines across your document portfolio.

Example Prompt: "Jane Doe is a Key Person who is leaving the firm. Analyse all my contractual obligations in light of this event."

3. Surface Precedent for Contract Drafting

Use Case: Locate the most relevant precedent and efficiently respond to detailed questionnaires by synthesising information from previously uploaded contracts, core documents, and policy materials.

Example Prompt: "Draft a comprehensive response to this questionnaire about our approach to balancing risk and return, including our philosophy and risk management processes."

‍

Prompting Tips

Troubleshooting

• Refresh the page when asking about new topics or sets of contracts: Your chat history is included in follow-up responses. We therefore recommend refreshing the browser to clear the chat when asking about separate topics or if you’ve changed the scope of contracts.
• Break down complex questions: If an answer seems incomplete it is likely trying to reference too many contracts and clauses at once. Ask a series of focused questions rather than one multi-part question.
• Verify with source documents: This is beta technology and we are continuing to improve accuracy and relevance of answers. Leverage clickable citations and

Specificity Leads to Better Results

• Narrow the scope of Documents: Leverage labels such as Contract Type and Group to reduce the number of documents to a smaller scope.
• Prompt with further specificity: Users can specify document scope even further, state the answer format (simple summary, clause comparison, risk assessment) and provide context on the circumstances for your search. "Analyse and Quote the termination rights in our Lease agreements with Alpha and Beta Company if I ([company]) is looking for early exit."
• Reference specific legal concepts over whole documents: While a prompt such as “Summarise Tax Clauses” may be helpful, more detailed prompts such as "Categorise all tax withholding provisions" are likely to provide more accurate and fulsome results.

‍