Sample data. Fictional companies. Nothing here is real.
GTMIQRead-only demo · Prototype
← Back to overview

One workflow, end to end: from buying signal to booked meeting

Sample data. Fictional companies. Nothing here is real.

Prototype. Every company, person and number on this page is fictional sample data.

Skills and tools

Demo only means I built the example to show the approach; it is not running in my live setup. Used in my sales career means hands-on user experience, not admin ownership. Learning / reference marks examples I wrote while building skills.

  • ApolloUsed in my live setup

    Step 1: sources ICP accounts and contacts; first-touch sequence enrollment.

  • ClayUsed in my live setup

    Step 2: waterfall enrichment for missing email, headcount and tech stack.

  • HubSpotUsed in my sales career

    Step 8: contact and deal fields written back, plus reporting. Hands-on user, not an admin.

  • AttioUsed in my live setup

    Step 8: optional CRM mirror of contacts and status.

  • SQL (Postgres)Learning / reference

    Step 3: stores accounts, signals and contacts; ranks accounts by fit. The snippet here is an example I wrote while building skills.

  • PythonLearning / reference

    Step 4: pandas cleanup, dedupe and top-account flags. Example I wrote while building skills.

  • MakeUsed in my live setup

    Step 5: orchestrates the steps with routing, retries and logging.

  • Claude APIUsed in my live setup

    Step 6: drafts openers and replies from account facts.

  • MCPUsed in my live setup

    Steps 3 and 6: exposes workspace data as tools the model can use (11 read, 4 write; writes need explicit authorization). Connected to Claude, ChatGPT, Grok, and Gemini.

  • MEDDICUsed in my live setup

    Steps 3 and 8: qualification fields feed scoring and deal records.

The 8 steps

  1. Step 1: Source

    ApolloUsed in my live setup

    Pull accounts and contacts that match the ICP filters.

    Input
    • Industry: logistics, healthcare, fintech, SaaS
    • Headcount: 50 to 1,500
    • Country: United States
    • Titles: IT Director, VP IT, CISO
    • Signal: hiring security roles (last 30 days)
    Output
    • Northwind Logistics · 420 emp.
    • Bluepeak Health · 1150 emp.
    • Copperline Fintech · 260 emp.
    • Harbor & Pine Legal · 90 emp.
    • Quillstone Software · 610 emp.
  2. Step 2: Enrich

    ClayUsed in my live setup

    Waterfall enrichment fills missing email, headcount and tech stack by trying providers in order until one returns a value.

    Input
    Quillstone Software · Marcus Lee, VP IT
    email: missing · tech stack: missing
    Output
    email: m.lee@quillstone.example (provider 2 of 3)
    headcount: 610 · stack: Okta, AWS, Jira
  3. Step 3: Store and score

    SQL · PostgresLearning / reference

    One query adds points for recent buying signals to each account's fit score and ranks the accounts.

    examples/score_accounts.sql
    -- Goal: rank accounts by fit plus recent buying signals. Works in Postgres and SQLite.
    WITH signal_points AS (
      SELECT account_id,
        CASE priority WHEN 'High' THEN 20 WHEN 'Medium' THEN 10 ELSE 5 END
          - CASE WHEN days_ago > 10 THEN 5 ELSE 0 END AS points
      FROM signals
    ),
    account_signals AS (
      SELECT account_id, SUM(points) AS signal_score, COUNT(*) AS signal_count
      FROM signal_points
      GROUP BY account_id
    ),
    scored AS (
      SELECT a.account_id, a.name, a.fit,
        COALESCE(s.signal_score, 0) AS signal_score,
        COALESCE(s.signal_count, 0) AS signal_count,
        a.fit + COALESCE(s.signal_score, 0) AS total_score
      FROM accounts a
      LEFT JOIN account_signals s ON s.account_id = a.account_id
    )
    SELECT RANK() OVER (ORDER BY total_score DESC) AS priority_rank,
      name, fit, signal_score, signal_count, total_score
    FROM scored
    ORDER BY priority_rank;
    priority_ranknamefitsignal_scoresignal_counttotal_score
    1Northwind Logistics82302112
    2Bluepeak Health7420194
    3Copperline Fintech6810178
    4Quillstone Software6110171
    5Harbor & Pine Legal550155

    Output of examples/run_sql.py on the fictional sample rows.

  4. Step 4: Clean and score in code

    PythonLearning / reference

    A short pandas pass normalizes emails, removes duplicates, tags seniority and flags target contacts.

    examples/clean_contacts.py
    import re
    import pandas as pd
    contacts = pd.read_csv("data/contacts_raw.csv")
    contacts["email"] = contacts["email"].str.strip().str.lower()
    before = len(contacts)
    contacts = contacts.drop_duplicates(subset="email", keep="first")
    print(f"Removed {before - len(contacts)} duplicate contact(s)")
    def seniority(title: str) -> str:
        t = title.lower()
        # \b is a word boundary, so "cto" does not match inside "director".
        if re.search(r"\b(chief|ciso|cio|cto|vp|head)\b", t):
            return "Executive"
        if re.search(r"\b(director|dir|manager)\b", t):
            return "Manager"
        return "Individual"
    contacts["seniority"] = contacts["title"].apply(seniority)
    contacts["target"] = contacts["seniority"].isin(["Executive", "Manager"]) & ~contacts["title"].str.contains("office", case=False)
    accounts = pd.read_csv("data/accounts.csv")
    out = contacts.merge(accounts[["account_id", "name", "fit"]], on="account_id").rename(columns={"name": "account"})
    out = out.sort_values(["fit", "seniority"], ascending=[False, True])
    print(out[["account", "first_name", "last_name", "title", "seniority", "target"]].to_string(index=False))
    Before (data/contacts_raw.csv)
    account_idfirst_namelast_nametitleemail
    1JordanAveryDirector of ITJORDAN.AVERY@northwind-sample.com
    1JordanAveryDir. ITjordan.avery@northwind-sample.com
    2PriyaLennoxCISOpriya.lennox@bluepeak-sample.com
    3MarcusOyelaranVP Engineeringmarcus.oyelaran@copperline-sample.com
    5ElenaFairbrookHead of Complianceelena.fairbrook@quillstone-sample.com
    5SamReyesIT Internsam.reyes@quillstone-sample.com
    4DanaWhitlockOffice Managerdana.whitlock@harborpine-sample.com
    After: Removed 1 duplicate contact(s)
    accountfirst_namelast_nametitlesenioritytarget
    Northwind LogisticsJordanAveryDirector of ITManagerTrue
    Bluepeak HealthPriyaLennoxCISOExecutiveTrue
    Copperline FintechMarcusOyelaranVP EngineeringExecutiveTrue
    Quillstone SoftwareElenaFairbrookHead of ComplianceExecutiveTrue
    Quillstone SoftwareSamReyesIT InternIndividualFalse
    Harbor & Pine LegalDanaWhitlockOffice ManagerManagerFalse

    Output of examples/clean_contacts.py on the fictional sample rows.

  5. Step 5: Orchestrate

    MakeUsed in my live setup

    A scenario runs the steps in order, routes by reply type, retries failed calls and logs every run.

    Trigger: schedule or new reply
    Router: new lead / reply
    Modules: enrich → score → draft
    Write to CRM + log
    Error handler: retry up to 3 times with backoff, then write the failed record and error to an activity log for review instead of dropping it.
  6. Step 6: Draft

    Claude APIUsed in my live setup

    The model gets only verified account facts, clear constraints and a list of banned claims, then returns a draft.

    prompt structure
    SYSTEM
    You write a first-touch email for a security services firm.
    
    ACCOUNT FACTS (only use these)
    - Company: Northwind Logistics, freight & logistics, ~420 employees
    - Signal: posted 3 security engineering roles this week
    - Contact: Dana Whitfield, Director of IT
    
    CONSTRAINTS
    - Under 90 words, plain text, one question at the end
    - Reference the signal once; no flattery
    - Offer a 20-minute call, no attachments
    
    BANNED CLAIMS
    - No guarantees, no "compliant in X days"
    - No customer names, no statistics, no pricing
    sample draft output
    Hi Dana,
    
    Saw Northwind is hiring three security engineers this week. Teams
    growing that fast often want an outside view of where the gaps are
    while new hires ramp up.
    
    We help mid-size logistics firms map that out without slowing the
    team down. Would a 20-minute call next week be useful?
    
    - Sample Sender
  7. Step 7: Approve

    Human step

    Replies are split by risk before anything goes out.

    Path A · Needs my approval
    Pricing, contract and meeting replies. I review and approve, edit or reject.
    Path B · Low-risk, sent automatically today
    First-touch sequence enrollment and simple replies such as out-of-office.
    Review queueMeeting requestBluepeak Health · R. Patel

    "Can we talk Thursday about scope and timeline?"

    Draft: "Thursday works. Would 2pm or 4pm ET suit you? I'll send an invite."

    Buttons shown for illustration; the demo is read-only.

  8. Step 8: Sync and report

    HubSpotUsed in my sales career

    Contact and deal fields are written back to the CRM, and a simple report rolls up the outcomes.

    objectfieldsample value
    Contactlifecycle_stageSales qualified lead
    Contactlast_outreach_statusReplied: meeting request
    DealdealstageDiscovery
    Dealmeddic_championIdentified
    Dealamount$72,000 (sample)
    segment (sample)reply ratemeetings bookedpipeline created
    Logistics9% (sample)3 (sample)$96k (sample)
    Healthcare7% (sample)2 (sample)$72k (sample)
    Fintech5% (sample)1 (sample)$30k (sample)

    All figures are fictional sample data, not results.