Backstory

Provides a strategic overview of the opportunity with AI-generated risks, next steps, and key topics. This tool analyzes the last 30 days of communication data and categorizes insights into: - Risks: Potential issues or concerns identified in conversations - Next Steps: Agreed-upon actions and commitments with the customer - Topics: Key themes and subjects being discussed This is the default tool for general questions about an opportunity. Use this tool unless the user explicitly asks for recent activity, detailed analysis, or company news. Args: peopleai_opportunity_id (int): People.ai opportunity ID to get the strategic summary for. Returns: text: Organized AI-generated summary showing current risks, committed next steps, and trending discussion topics from recent communications.

Retrieves the last 30 days of communication activity between your team and the account. This includes AI-generated summaries of emails and meeting transcripts with key facts extracted and classified by importance (strategic, tactical, high-risk). Noise and small talk are filtered out to highlight revenue-relevant discussions and potential risks. Use this only when the user asks specifically about recent activity or communications. Do NOT use for general account questions. Args: peopleai_account_id (int): People.ai account ID to retrieve communication history for. Returns: text: A formatted AI-generated summary of recent activity with extracted facts, risk indicators, and strategic insights from the last 30 days.

Retrieves the last 30 days of communication activity between your team and the opportunity. Use this tool when asking for details about an opportunity. It provides a summarized version of the communication data that is also used by the opportunity_status tool. The AI-generated summary includes key facts extracted from emails and meeting transcripts, classified by importance (strategic, tactical, high-risk). Noise and small talk are filtered out to highlight revenue-relevant discussions and potential risks. Use this only when the user asks specifically about recent activity or communications. Do NOT use for general opportunity questions. Args: peopleai_opportunity_id (int): People.ai opportunity ID to retrieve communication history for. Returns: text: Formatted AI-generated summary of recent email and meeting activity with extracted facts, risk indicators, and strategic insights from the last 30 days.

Retrieves the scorecard for a specific account, including all configured questions and their current answers. The scorecard provides a structured assessment of account health based on predefined qualification or evaluation criteria. Each item includes the question text and the most recent answer. If no scorecard is returned, either: - No scorecard has been configured for this record, or - Scorecard ingestion has not been set up in the environment Use this when the user asks about scorecard status, account plans, or evaluation progress for an account. For an opportunity's scorecard, use get_opportunity_scorecard instead. Do NOT use for general account questions.

First step for any request for a LIST or a COUNT of Backstory accounts or opportunities described in plain language — e.g. 'accounts I own with no upcoming meetings', 'open opps closing this quarter', "accounts owned by Jane Doe", "my team's opportunities". Reports how many records match and how the request was understood; returns NO records. The interpretation and the query behind the count are AI-generated; the count itself comes from People.ai data. Relay the count and the interpretation to the user and then STOP — wait for their answer. If exceeds_max is true, tell them the match set is larger than the 1,000-record maximum before they decide. If they add a condition or correct the reading, call preview_records AGAIN with the full request including that change, and show the new count. Only when they confirm, or ask for a subset, call fetch_records with the same request. Pass the user's wording as-is; do not translate it into field names. If the response contains owner_disambiguation instead of a count, several people match the owner name: relay the candidates (name and id) to the user, let THEM choose — never pick one yourself — then call preview_records again with the chosen person folded into the query text as "owned by <Name> (id: <id>)". Calling fetch_records with the ambiguous name returns the same candidates; it never guesses.

Analyzes a sales opportunity's or account's current situation and finds similar historical cases with resolution outcomes. The extracted situation and each candidate's similarity and resolution analysis are AI-generated. Given an opportunity (or an account) and a description of the challenge or concern, this tool: 1. Analyzes recent deal activity to understand the specific situation 2. Finds past opportunities or accounts that faced similar challenges 3. Evaluates how similar each historical case is using a structured question checklist 4. Reports how those situations were resolved — what actions were taken and whether they succeeded When given an opportunity, it analyzes that single deal. When given an account, it analyzes the account's activities. In both cases the similar cases returned are past opportunities or accounts, and records tied to the source are excluded: when searching by account, the account itself and all of its opportunities; when searching by opportunity, the opportunity itself and its parent account. Returns up to 4 most similar cases (those scoring above 70% similarity) with full resolution analysis. If no candidates meet the threshold, reports that no similar situations were found. Use this tool when the user wants to: - Understand how similar deal situations were handled in the past - Learn what actions worked (or didn't) in comparable scenarios - Get actionable precedents for a current deal or account challenge - Seek advice or consider options for overcoming a specific challenge or obstacle in their deal or account Args: query (str): A concise description of the deal challenge or concern to investigate. Example: "Buyer is pushing back on pricing and evaluating competitor alternatives" object_type (str): The type of record to analyze — "opportunity" for a single deal, "account" for the account's situation across all of its activity and deals. peopleai_object_id (int): The internal People.ai ID of the record to analyze — an account ID or an opportunity ID, matching object_type. Use find_account or find_record_by_crm_id to obtain this. Returns: dict with: - situation: AI-generated extracted situation description (anonymized with buyer/seller roles) - candidates: Up to 4 most similar historical cases (above 70% match). Each candidate has an object_type of "opportunity" or "account". The opportunity_* fields are populated only for "opportunity" candidates and are null for "account" candidates. Fields: - object_type: "opportunity" or "account". - account_id, account_name: the account — itself for an "account" candidate, or the account the opportunity belongs to for an "opportunity" candidate - situation_summary: AI-generated summary of what happened in this deal/account regarding the target situation - match_score: AI-generated similarity score (0.0 - 1.0) - resolution_status: AI-generated status — "resolved", "partially_resolved", or "unresolved" - resolution_explanation: AI-generated explanation of why this status was assigned - actions_taken: AI-generated key actions the seller took to address the situation - outcome: AI-generated description of what resulted from those actions - opportunity_* (opportunity candidates only): opportunity_id, opportunity_name, opportunity_crm_id, opportunity_owner_name, opportunity_amount, opportunity_close_date, opportunity_is_closed, opportunity_is_won

Retrieves the most relevant accounts and opportunities to the current user's profile. This tool returns top active accounts and opportunities that may be of interest for the user, applying filters and sorts based on the user's role or a custom profile configuration. This is not an exhaustive list of all owned or visible accounts and opportunities. IMPORTANT USAGE RULES: 1. Call this tool in one of the two cases: User does not specify a particular account name or CRM ID. Then present a summary of the accounts and opportunities to the user, requesting further specification. User asked about an account, but find_account tool didn't find it by name. Then use this tool to check its response for possible misspellings of the account name 2. Do not call this tool more than once per interaction, use its output from your context if needed. 3. DO NOT automatically call other tools in a loop for all of the accounts/opportunities retrieved. Wait for the user to specify which record(s) they want to explore further. 4. This tool provides basic reference information about accounts and opportunities. Use downstream tools (status, activity, news, etc.) to collect detailed information for analysis after the user explicitly selects one or more specific accounts or opportunities. If the desired account or opportunity is not found in the response, use: - find_account: to search by account name - find_record_by_crm_id: to look up by CRM ID Returns: list[dict]: Array of accounts with their open opportunities. Only the 20 most relevant accounts are returned. Before using any account-specific tools, you MUST show the list of accounts to the user and let them choose one or more.