AI

Reference · 10 stacks

Workflow Stacks

API combinations that unlock multiple use cases — with example data flows and missing-tool notes.

API CombinationBeginner

Booking Operations Stack

Wire Calendly, Gmail, Slack, and OpenAI together for real-time booking notifications, lead qualification, and appointment follow-up.

Example workflow

Calendly webhook fires on new booking. OpenAI reads the invitee's custom question answers and drafts a qualification summary. Slack notification posts to #bookings with client name, service type, and prep notes. 24 hours before the appointment, Gmail sends a confirmation with prep instructions generated from CRM notes.

Business value

Staff arrive prepared for every appointment. No-shows drop when confirmation sequences include personalized prep content.

Recommended for

Coaches and consultantsMedical and dental practicesLegal and financial advisorsHome service providers

Risks

  • Webhook reliability depends on Calendly plan tier
  • Slack notifications with client PII should use private channels
  • Lead qualification AI may misjudge intent — use as advisory only
API CombinationBeginner

Email + Calendar Intelligence Stack

Connect Gmail, Google Calendar, and OpenAI to automate inbox triage, meeting prep, and follow-up reminders — the most common starting stack for service businesses.

Example workflow

Every morning at 7 AM, fetch unread Gmail messages and today's calendar events. Send both to OpenAI with a prompt to produce a prioritized briefing: urgent emails first, then meeting prep notes with context from related threads. Deliver the briefing as a Gmail draft or Slack message.

Business value

Saves 30–45 minutes of morning admin daily. Owners start the day knowing exactly what needs attention instead of scanning inboxes reactively.

Recommended for

ConsultantsAgenciesProfessional servicesSolo operators

Risks

  • AI may miss subtle urgency cues — always review before acting
  • OAuth token refresh must be handled for unattended runs
  • Calendar + email data contains client PII
API CombinationIntermediate

HubSpot Sales Execution Stack

HubSpot as CRM source of truth, Gmail for email context, and Claude/OpenAI for briefings, nurture sequences, proposal follow-ups, and pipeline risk detection — the daily sales ops stack.

Example workflow

Daily morning: pull today's calendar meetings from Google Calendar. For each external meeting, HubSpot provides deal context and Gmail provides last 3 threads. Claude generates a one-page sales briefing to Slack 1 hour before the call. Separately, a daily job scans deals in 'Proposal Sent' stage with no reply in 5+ days — OpenAI drafts follow-up emails as Gmail drafts. Weekly: pipeline risk scan flags deals with 14+ days no activity and posts rep-specific cleanup lists to Slack.

Business value

Unifies prep, follow-up, and risk detection in one CRM-centric workflow. Reps spend less time in tabs and more time on conversations. Proposal follow-up alone often recovers 10–15% of stalled deals.

Recommended for

HubSpot-centric SMB sales teamsAgencies with 5–20 person sales teamsFounder-led sales with heavy email volume

Risks

  • Gmail API scope approval requires Google security review for production
  • Multiple daily automations can overwhelm reps with Slack noise — batch and prioritize
  • AI briefings with wrong CRM data erode trust — validate data quality first
API CombinationIntermediate

Knowledge Base Builder Stack

Feed existing documents from Google Drive into OpenAI, then publish structured SOPs and training content to Notion — turns tribal knowledge into onboarding assets.

Example workflow

Owner uploads a recorded process video transcript and existing checklist PDF to a Drive folder. OpenAI converts both into a step-by-step SOP with roles, tools needed, and common mistakes. The SOP publishes to a Notion database. Staff receive a Slack link to the new page. Quarterly: OpenAI diffs updated Drive docs against Notion pages and flags outdated SOPs.

Business value

New hires onboard in days instead of weeks. SOPs stay current without the owner rewriting docs manually. Reduces repeated training questions by 40–60%.

Recommended for

FranchisesMulti-location service businessesTeams with high turnoverGrowing companies formalizing processes

Risks

  • AI-generated SOPs may miss safety-critical steps — require expert review
  • Drive folder permissions must exclude personal files
  • Notion page structure changes can break automated publishing
API CombinationIntermediate

Local Lead Discovery & CRM Stack

Combine Google Places and Yelp for local B2B prospecting, score and personalize with OpenAI, and sync clean leads into HubSpot — the core stack for outbound local sales teams.

Example workflow

Weekly: search Google Places by ICP category and city radius. Cross-match with Yelp for reviews and ratings. OpenAI scores each business for fit and tags pain signals from review text. Deduplicate against HubSpot companies by domain and phone. Create new contacts with score, pain tags, and a draft outreach angle in a custom property. Rep reviews top-scored leads and sends personalized emails from Gmail drafts.

Business value

Replaces 5–10 hours of manual list building per week. Pain-signal tagging improves reply rates vs. generic cold email. CRM dedup prevents duplicate outreach to the same account.

Recommended for

B2B agencies selling to local businessesSaaS with local SMB ICPProfessional services firms doing territorial prospecting

Risks

  • Google Places and Yelp rate limits require batching for large territories
  • Cold outreach compliance varies by region — include opt-out and accurate sender info
  • Yelp data may not cover all B2B verticals — supplement with Outscraper for gaps
API CombinationIntermediate

Local Reputation Intelligence Stack

Pull reviews from Google Business Profile and Yelp, analyze with OpenAI, and generate responses or content — built for local businesses managing online reputation.

Example workflow

Weekly: fetch new Google and Yelp reviews. OpenAI clusters them into themes (service quality, pricing, wait times) and drafts owner responses for negative reviews. Positive review language is extracted for social content ideas. Owner approves responses before publishing via Google Business Profile API.

Business value

Faster review response times improve local SEO rankings. Theme analysis reveals operational fixes competitors aren't making.

Recommended for

RestaurantsSalons and spasHome servicesRetail shops

Risks

  • Auto-published review responses can sound generic — require human approval
  • Yelp API rate limits may require batching
  • Competitor data usage must comply with platform terms
API CombinationIntermediate

Revenue Recovery Stack

Combine Square POS data, Gmail, Twilio SMS, and OpenAI to win back inactive customers and recover revenue from no-shows and lapsed buyers.

Example workflow

Monthly: Square API pulls customers with no purchase in 60+ days. OpenAI drafts personalized winback emails referencing their last purchase. High-value lapsed customers also receive an SMS via Twilio. Track responses in a Google Sheet and log conversions back to Square customer notes.

Business value

Reactivating 5–10% of lapsed customers often pays for the entire automation stack. No-show recovery alone can recover 15–20% of lost appointment revenue.

Recommended for

Retail storesRestaurantsFitness studiosSalons

Risks

  • SMS requires documented opt-in — verify before sending
  • Over-messaging lapsed customers damages brand trust
  • Personalization based on purchase history must respect privacy policies
API CombinationAdvanced

Salesforce Enterprise Pipeline Stack

Salesforce for CRM and pipeline management, Microsoft Outlook for email context, and Claude/OpenAI for hygiene monitoring, risk detection, briefings, and lead scoring — built for Microsoft-centric sales orgs.

Example workflow

Weekly: Salesforce API exports full pipeline. Rule engine flags duplicates, stale stages, and missing next steps. Claude generates a rep-by-rep hygiene digest posted to Slack. Daily: active opportunities scanned for risk — stage age, Outlook email response gaps, missing tasks. OpenAI writes risk narrative and recovery suggestions. Before external meetings: pull Salesforce account + Outlook thread history → Claude briefing to rep's inbox. New inbound leads scored via enrichment + AI and written to Salesforce lead score field.

Business value

Brings AI intelligence to Salesforce without Einstein licensing. Hygiene and risk detection improve forecast accuracy. Outlook integration matches how enterprise reps actually work.

Recommended for

Salesforce-centric B2B teamsMicrosoft 365 organizationsRevOps teams managing 50+ opportunity pipelines

Risks

  • Salesforce API limits and complex object models require careful integration design
  • Microsoft Graph permissions need IT admin approval in enterprise environments
  • Automated CRM writes must respect validation rules and trigger side effects
API CombinationBeginner

Scrape-to-Sheet Research Stack

Use Outscraper or Apify to bulk-extract business listings and reviews into Google Sheets, then run Claude or OpenAI for lead scoring, account research, and review-based pain signal detection.

Example workflow

On demand: Outscraper pulls 500 Google Maps businesses matching 'property management' in Dallas with reviews. Apify scrapes each business website homepage for service keywords. Data lands in Google Sheets. Claude batch-processes review text to extract pain themes (slow response, billing complaints, staffing). OpenAI scores ICP fit and writes a one-paragraph account summary per row. Rep filters top 50 and imports to CRM or outreach tool.

Business value

Low-code path to bulk prospecting without building a custom scraper. Sheets becomes the review layer before CRM commit. Review pain themes directly fuel personalized outreach.

Recommended for

Small sales teams without engineering resourcesAgencies building vertical-specific prospect listsMarketers doing ABM list research before campaigns

Risks

  • Scraping volume costs add up — budget per territory before bulk runs
  • Sheet-based workflows don't scale past ~5,000 rows without performance issues
  • Scraped data freshness degrades — re-run quarterly for active territories
API CombinationBeginner

Spreadsheet Operations Stack

Use Google Sheets as your operational database, Gmail for communication, and OpenAI for summarization — the lowest-friction stack for spreadsheet-driven SMBs.

Example workflow

Website form submissions append rows to a Quotes tracker sheet via Zapier. OpenAI reads new rows daily, drafts follow-up emails for quotes older than 3 days, and flags overdue items. Friday: OpenAI reads the full sheet and produces a weekly owner briefing with pipeline status, overdue follow-ups, and revenue forecast.

Business value

Turns a spreadsheet you already use into an active workflow engine without adopting new software. Weekly briefings replace manual status meetings.

Recommended for

ContractorsFreelancersSmall agenciesAny business already running on Google Sheets

Risks

  • Spreadsheet formulas can break if AI writes to wrong cells — use append-only patterns
  • Concurrent edits by humans and automation cause conflicts
  • Sheet access permissions must be scoped to service account only