Skip to content

[Agent Idea]: Sales Call News Researcher — Scrape Recent News About a Company Before a Call #4651

@9series-engineer

Description

@9series-engineer

Problem Statement

Sales teams spend significant time each morning manually researching the companies they're scheduled to meet with. This involves opening multiple browser tabs, searching for recent news, scanning articles, and taking notes — often for 5–10 companies per day. This repetitive, time-consuming process leads to inconsistent preparation, missed insights, and wasted hours that could be spent on higher-value activities like building relationships and closing deals.

Proposed Solution

Build a Sales Call News Researcher agent that automatically prepares personalized company news briefings each morning before your calls. The agent runs a daily workflow that connects your calendar to web search and email delivery.

How It Works

  1. Scans Your Calendar: Each morning, the agent reviews your Google Calendar for any scheduled meetings or calls with external companies for that day.
  2. Identifies Companies: It extracts company names from meeting titles, descriptions, and attendee email domains (e.g., @acme.com → Acme Inc.).
  3. Fetches Latest News: For each identified company, it searches the web for the most recent and relevant news articles — including funding rounds, product launches, executive changes, earnings reports, partnerships, and press coverage.
  4. Delivers Insights: You receive personalized emails via Gmail, each dedicated to a specific company you're meeting with that day. Each email contains a curated list of news headlines, brief descriptions, and direct links to the full articles.

Example Output (Email)

Subject: 📰 Pre-Call Briefing: Acme Corp — 3 Recent News Items

Hi [User],

You have a meeting with Acme Corp today at 2:00 PM. Here's what's been happening:

1. "Acme Corp Raises $50M Series C to Expand AI Platform"
   - TechCrunch, Feb 10, 2026
   - Acme closed a $50M round led by Sequoia to scale their enterprise AI offering.
   - [Read Full Article →]

2. "Acme Corp Partners with Salesforce for CRM Integration"
   - Business Wire, Feb 8, 2026
   - New strategic partnership to embed Acme's analytics into Salesforce dashboards.
   - [Read Full Article →]

3. "Acme Corp CEO Interviewed on Future of B2B SaaS"
   - Forbes, Feb 5, 2026
   - CEO Jane Smith discussed market trends and Acme's 2026 growth strategy.
   - [Read Full Article →]

Good luck with your call!

Use Case

This agent is designed for sales teams, account executives, business development reps, and anyone who regularly takes external meetings. Specifically:

  • Pre-call preparation: Walk into every meeting informed about your prospect's latest developments.
  • Account management: Stay updated on existing clients without manual effort.
  • Partnership discussions: Know your counterpart's recent moves before negotiating.
  • Investor relations: Be aware of portfolio company news before check-ins.

Agent Architecture

Nodes

Node Type Description
calendar_scanner Function Node Connects to Google Calendar API, retrieves today's meetings, extracts company names and meeting metadata
company_identifier LLM Node Takes raw meeting data and intelligently identifies the external company name (handles edge cases like internal meetings, personal events, ambiguous titles)
news_fetcher Function Node Uses web search tool to find recent news articles for each identified company (last 7–30 days)
news_curator LLM Node Filters, ranks, and summarizes the most relevant articles; removes duplicates and irrelevant results
email_composer LLM Node Generates a clean, scannable email briefing for each company
email_sender Function Node Sends the composed email via Gmail API

Edges (Flow)

calendar_scanner → company_identifier → news_fetcher → news_curator → email_composer → email_sender

Integrations Required

Integration Purpose Auth Method
Google Calendar API Read today's scheduled meetings OAuth 2.0
Web Search Tool Fetch recent news articles for each company API Key (already available in Hive tools)
Web Scraper Tool Extract article content for summarization Already available in Hive tools
Gmail API Send personalized briefing emails OAuth 2.0

Self-Improvement Opportunities

This agent fits well with Hive's self-improving architecture:

  • When news results are irrelevant: The agent learns which types of search queries yield better results for different company types (startup vs. enterprise, public vs. private).
  • When company identification fails: If the agent misidentifies or misses a company from a calendar event, it adapts its extraction logic based on feedback.
  • When emails go unread: If the user consistently ignores briefings for certain meeting types, the agent can learn to skip those (e.g., internal 1:1s that were misclassified).
  • When article quality is low: The agent can evolve to prefer certain news sources over others based on user engagement patterns.

Human-in-the-Loop

  • Optional review step: Before emails are sent, the agent can pause and present the draft briefings for approval (useful during initial setup).
  • Feedback mechanism: Users can rate briefings (thumbs up/down) to help the agent improve over time.
  • Configuration: Users can set preferences for news recency window, number of articles per company, and preferred news sources.

Additional Context

  • This aligns with the Sales Agent template mentioned in the Hive roadmap.
  • Could be extended to support Slack delivery (in addition to email) using Hive's existing Slack integration.
  • Could be extended to include LinkedIn profile insights for individual attendees.
  • The web search and web scraper tools are already available in Hive's MCP tools package.

Metadata

Metadata

Assignees

No one assigned

    Labels

    enhancementNew feature or requesthelp wantedExtra attention is needed

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions