serenorg/job-seeker

Seren Job Seeker

AI-powered job search agent following Taariq Lewis's proven strategy: network with hiring managers directly, skip the ATS black hole.

⚠️ IMPORTANT DISCLAIMERS

READ THIS BEFORE USING

Ethical Use Only

• This skill is for legitimate job seeking and professional networking
• Respect privacy: Only contact people with professional context
• Follow platform ToS: Don't scrape data in violation of terms of service
• No spam: Personalized outreach only, never mass unsolicited emails
• Respect opt-outs: Honor requests to stop contact immediately

Data Privacy

• Contact data (emails, phone numbers) is personal information
• Handle responsibly and delete when no longer needed
• Comply with GDPR, CAN-SPAM, and local data protection laws
• Never sell or share contact information

No Guarantees

• This skill does not guarantee job offers
• Success depends on your skills, market conditions, and approach
• Networking takes time — expect weeks to months, not days

Publisher Costs

• Apollo.io: $0.04 per contact lookup (verified emails)
• AlphaGrowth: $0.03 per company discovery + $0.01 per email verification
• Perplexity + Exa: $0.22 per company research
• Playwright: $0.04 per event discovery
• Seren Models (GPT-5.2): $3.00 per outreach generation
• Total: $11-18 per comprehensive search (50 companies → 10 targets → 3 outreach emails)

---

When to Use This Skill

Activate this skill when the user mentions:

• "help me find a job"
• "search for [role] positions at [companies]"
• "find hiring managers at [company]"
• "generate personalized outreach emails"
• "discover networking events in [location/industry]"

For Claude: How to Invoke This Skill

When the user asks to find jobs or apply to companies, follow this 7-phase workflow:

Phase 0: User Profile Extraction

Goal: Extract user's background from resume + LinkedIn data export BEFORE searching for jobs

CRITICAL: Run this FIRST. Without user context, we can't filter companies, tailor outreach, or auto-fill applications.

BOTH resume AND LinkedIn export are REQUIRED.

Step 1: User downloads LinkedIn data export (REQUIRED)

How to download LinkedIn data export:

1. Go to linkedin.com/mypreferences/d/download-my-data
2. Select "Download larger data archive" → Check all data types
3. Click "Request archive" (takes 10-15 minutes)
4. LinkedIn emails download link
5. Download ZIP file (e.g., Basic_LinkedInDataExport.zip)

python3 scripts/agent.py extract-profile \
  --resume resume.pdf \
  --linkedin-export linkedin-export.zip \
  --output user_profile.json

What gets extracted:

From Resume:

• Work history: Companies, roles, duration, achievements
• Skills & tech stack: Languages, frameworks, tools
• Education: Degrees, schools, years
• Notable achievements: Quantified impact

From LinkedIn Export (REQUIRED):

• Connections: Mutual connections at target companies
• Recommendations: Who endorsed your skills
• Activity: Recent posts, comments (shows current interests)
• Complete work history: Including descriptions LinkedIn has
• Skills endorsements: Validation from colleagues

Tools used:

• GPT-5.2: Parse resume PDF/DOCX + LinkedIn CSV/JSON → structured JSON
• No scraping: User provides data directly (ToS-compliant)

Cost: $0.50 per user (resume + LinkedIn parsing)

Present to user:

Profile extracted!

Alex Chen - Senior ML Engineer
• 8 years experience
• Skills: Python, Rust, PyTorch, distributed training
• Location: SF, NYC, or Remote
• Salary: $180k+ minimum
• Top achievement: Reduced training time 40%
• LinkedIn connections: 487 (23 at target companies)

Ready to find companies matching your profile?

---

Phase 1: Company Discovery

Goal: Find 50 target companies matching user criteria (using Phase 0 profile)

Uses Phase 0 profile to filter:

• Match seniority (don't suggest junior roles to seniors)
• Match tech stack (PyTorch experience → ML companies)
• Match location preferences (remote, SF, NYC)
• Match industry experience (fintech background → fintech startups)

# Discover companies via AlphaGrowth
# Example: Find 50 AI startups in SF with 10-100 employees
python3 scripts/agent.py discover \
  --profile user_profile.json \
  --role "Senior ML Engineer" \
  --industry "AI" \
  --location "SF" \
  --limit 50

What this does:

• Searches AlphaGrowth database for companies matching filters
• Returns: company name, domain, size, funding, location
• Cost: $0.03 per company discovered (~$1.50 for 50 companies)

Present to user:

Found 50 companies matching your criteria:

Top 10:
1. Anthropic (anthropic.com) - 150 employees, Series C, SF
2. Runway ML (runwayml.com) - 85 employees, Series B, NYC
3. Cohere (cohere.ai) - 120 employees, Series C, Toronto
...

---

Phase 2: Company Research

Goal: Deep research on top 20 companies (culture, tech stack, recent news, hiring signals)

# For each company:
# 1. Perplexity: "Research [company] culture, tech stack, recent news, and hiring signals"
# 2. Exa: Semantic search for engineering blogs, job postings, employee interviews

What this does:

• Uses Perplexity to research company background, culture, recent developments
• Uses Exa to find technical content (engineering blogs, talks, open source)
• Identifies hiring signals (recent funding, team growth, new products)
• Cost: $0.22 per company (~$4.40 for 20 companies)

Present to user:

Researched 20 companies. Top insights:

Anthropic:
  • Culture: Research-focused, safety-first, academic feel
  • Tech: Rust, Python, PyTorch, distributed systems
  • Recent: Just launched Claude 4.6, hiring ML engineers
  • Signals: 3 job postings this week, 15% team growth

---

Phase 3: Hiring Manager Discovery (Apollo Primary)

Goal: Find decision-makers (hiring managers, team leads, VPs of Engineering)

Primary Tool: Apollo.io

# Apollo people search with filters:
# - Company: [target company domain]
# - Titles: "Engineering Manager", "VP Engineering", "Director of Engineering", "Head of ML"
# - Seniority: Manager, Director, VP, C-Suite
# Returns: Name, title, verified email, LinkedIn, phone (optional)

Why Apollo is better:

• ✅ Verified emails (90%+ accuracy) vs scraped emails
• ✅ Job change tracking — alerts when people switch companies
• ✅ Hiring signals — shows when companies are actively hiring
• ✅ ToS-compliant — Apollo provides data legally
• ✅ Cost-effective — $0.04 per contact vs $0.25/company for Apify scraping

Secondary Tool: Playwright (LinkedIn context)

# For top 3-5 targets per company:
# 1. Visit LinkedIn profile (public data only)
# 2. Extract: Recent posts, conference talks, interests, mutual connections
# 3. Gather social context for personalization

What this does:

• Apollo returns 5-10 hiring managers per company with verified emails
• Playwright enriches top candidates with social context (posts, talks, interests)
• Identifies warm intro opportunities (mutual connections)
• Cost: $4.00 per company batch (~$40 for 10 companies)

Present to user:

Found 47 hiring managers across 10 companies:

Anthropic (5 contacts):
  1. Sarah Chen - VP Engineering
     • Email: [email protected] (verified ✓)
     • LinkedIn: Recent post about scaling ML training
     • Context: Spoke at MLSys 2026, interested in distributed systems

  2. Michael Rodriguez - Engineering Manager, Safety Team
     • Email: [email protected] (verified ✓)
     • Mutual: 2 connections (Jane Doe, John Smith)

---

Phase 4: Event Discovery

Goal: Find networking events, conferences, meetups where hiring managers will be

# Exa semantic search:
# "AI conferences in San Francisco March 2026"
# "Engineering meetups at Anthropic headquarters"

# Playwright scraping:
# - Eventbrite: AI/ML events
# - Meetup.com: Tech meetups
# - Luma: Startup events
# - LinkedIn Events: Company-hosted events

What this does:

• Discovers upcoming events in target location/industry
• Identifies which hiring managers are attending or speaking
• Provides context for in-person networking opportunities
• Cost: $0.04 per event discovery (~$0.40 for 10 events)

Present to user:

Found 8 networking opportunities:

1. AI Safety Summit - March 15, 2026 - SF
   • Speakers: Sarah Chen (Anthropic), 3 other targets
   • Registration: $50 early bird

2. Bay Area ML Meetup - March 22, 2026 - SF
   • Host: Anthropic (office tour + tech talk)
   • Free, RSVP required

---

Phase 5: Email Verification & Personalized Outreach

Goal: Verify emails, generate personalized outreach that demonstrates value

Email Verification (AlphaGrowth)

# For each contact email from Apollo:
# AlphaGrowth email verification API
# Returns: deliverable, risky, or invalid
# Only proceed with "deliverable" emails

Why verify emails:

• ✅ Reduces bounce rate — avoid damaging sender reputation
• ✅ Improves deliverability — ISPs trust senders with low bounce rates
• ✅ Saves costs — don't waste outreach on dead emails
• ✅ Professional — bounced emails look careless
• Cost: $0.01 per email verified (~$0.50 for 50 contacts)

Outreach Generation (Seren Models - GPT-5.2)

# For each verified contact, generate personalized email:
# Input:
# - Hiring manager name, title, company
# - Company research (culture, tech stack, recent news)
# - Social context (recent posts, talks, interests)
# - Mutual connections
# - User's background and value proposition
# - Event opportunity (if applicable)
# - **Application ID from Phase 7** (if using double-tap strategy)

# Output: 3-paragraph email
# 1. Personal hook (reference their work, post, or talk)
# 2. Application reference (if double-tap) + Value proposition
# 3. Soft ask (coffee, event meetup, or intro call)

Outreach Principles (from Taariq's blog):

• Quality over quantity — 3 great emails > 30 generic ones
• Value-first — lead with what you can offer, not what you need
• Research-backed — reference specific details (posts, projects, tech)
• No resume dumps — skip "attached is my resume", focus on conversation
• Event-based — "I'll be at AI Safety Summit, would love to chat" (warmer than cold email)
• Double-tap — Reference your ATS application to make it actionable

Cost: $3.00 per outreach email (~$9.00 for 3 high-priority emails)

Present to user:

Email verification complete:
  • 47 emails checked
  • 43 deliverable (91%)
  • 3 risky (skipped)
  • 1 invalid (skipped)

Generated 3 personalized emails for top targets:

---

Subject: Your MLSys talk on distributed training + my application (#AN-2026-00142)

Hi Sarah,

I just applied for the Senior ML Engineer - Safety Team role (Application
#AN-2026-00142) and wanted to reach out directly.

I caught your MLSys 2026 talk on scaling distributed training to 10k GPUs — the
part about gradient compression was brilliant. I've been working on similar
challenges at [Current Company], where we reduced training time 40% by
implementing a custom allreduce algorithm.

I'm excited about Anthropic's approach to safety-constrained training at scale
and would love to learn more. I'll be at the AI Safety Summit on March 15 —
any chance you're free for coffee that morning?

Looking forward to connecting,
[Your Name]

---

---

Phase 6: Application Tracking

Goal: Track outreach, responses, interviews, and follow-ups

# Store in structured format (SQLite or Google Sheets):
# - Company name
# - Contact name, title, email
# - Outreach sent date
# - Response received (yes/no)
# - Interview scheduled (yes/no)
# - Status (pending, interview, offer, rejected)
# - Follow-up date
# - Notes

What this does:

• Maintains CRM-style database of all outreach
• Tracks response rates and conversion funnel
• Schedules follow-ups (if no response in 7 days)
• Measures effectiveness (which companies/approaches work best)

Present to user:

Application Tracker Summary:

Outreach sent: 43 emails
Responses: 8 (18.6%)
Interviews scheduled: 3 (7.0%)
Offers: 0
Pending follow-ups: 12 (due this week)

Top performers:
  • Event-based outreach: 4/10 responses (40%)
  • Mutual connection intro: 2/5 responses (40%)
  • Cold email (research-backed): 2/28 responses (7%)

---

Phase 7: Automated Job Applications (Double-Tap Strategy)

Goal: Apply via ATS to the SAME 10 companies where you're networking, then reference your application in hiring manager outreach

CRITICAL: Phase 7 operates on the SAME companies from Phase 3, not a separate list.

The Double-Tap Strategy:

1. Apply via ATS (Phase 7) → Get confirmation number
2. Email hiring manager (Phase 5) → Reference application ID
3. Result: Your application doesn't go to a black hole + hiring manager knows to look for it

Why this works:

• ✅ Double visibility: HR sees application, hiring manager sees email
• ✅ Shows initiative: Applied formally AND reached out directly
• ✅ Actionable: Hiring manager can pull your application by ID
• ✅ Higher conversion: 10-15% (vs 2-5% ATS alone, 7-40% networking alone)

Example flow for Anthropic:

# 1. Apply via ATS (Phase 7)
→ Submit application
→ Get confirmation: "Application #AN-2026-00142"

# 2. Email Sarah Chen (Phase 5)
→ Subject: "Your MLSys talk + my application (#AN-2026-00142)"
→ Body: "I just applied for Senior ML Engineer (Application #AN-2026-00142)
         and wanted to reach out directly. I caught your MLSys talk..."

Workflow per company:

# For each of the 10 target companies:
# 1. Scrape careers page for matching job postings
# 2. Apply via ATS (auto-fill, submit, get confirmation ID)
# 3. Generate networking email that references application ID
# 4. User sends networking email

What this does:

• Job discovery: Playwright scrapes company careers pages
  • Supports: Greenhouse, Lever, Workday, Ashby, custom ATS
  • Filters by role keywords, location, seniority
• Resume tailoring: GPT-5.2 generates role-specific resume variants
  • Highlights relevant experience for each position
  • Optimizes for ATS keyword matching
• Cover letter generation: Personalized per role
  • References company research from Phase 2
  • Explains fit for specific position
• Form automation: Playwright fills forms automatically
  • Personal info (name, email, phone, LinkedIn)
  • Work history (auto-populated from master resume)
  • Education, skills, portfolio links
  • Handles dropdowns, checkboxes, text fields
• Document upload: Attaches PDF resume + cover letter
• CAPTCHA solving: Uses 2Captcha for bot detection bypass
• Submission: Clicks submit, saves confirmation

Supported ATS platforms:

• ✅ Greenhouse (60% of startups)
• ✅ Lever (25% of startups)
• ✅ Workday (enterprise companies)
• ✅ Ashby (newer startups)
• ✅ Custom forms (best-effort)

Cost: $2-5 per company

• Job discovery: $0.10 (scrape careers page)
• Resume tailoring: $0.50 per variant (GPT-5.2)
• Cover letter: $0.50 per role (GPT-5.2)
• Form automation: $1.00 per application (Playwright)
• CAPTCHA solving: $0.50 per captcha (2Captcha)

Present to user:

Found 12 job postings across 10 companies:

Anthropic (2 openings):
  1. Senior ML Engineer - Safety Team
     • Location: SF (Hybrid)
     • Salary: $180k-250k
     • Applied: ✓ (submitted 2026-03-10 14:35 UTC)
     • Confirmation: Application #AN-2026-00142

  2. Staff Engineer - Infrastructure
     • Location: Remote (US)
     • Salary: $200k-280k
     • Applied: ✓ (submitted 2026-03-10 14:38 UTC)
     • Confirmation: Application #AN-2026-00143

Runway ML (1 opening):
  1. Senior Software Engineer - ML Platform
     • Location: NYC (Onsite)
     • Salary: $160k-220k
     • Applied: ✓ (submitted 2026-03-10 14:42 UTC)
     • Confirmation: Application #RML-2026-00089

Summary:
  • Jobs found: 12
  • Applied: 12 (100%)
  • Failed: 0
  • Total cost: $36.00 (12 applications × $3.00 avg)
  • Time saved: ~6 hours (vs manual application)

Best practices:

• Apply selectively: Don't spam every posting
• Tailor per role: Generic applications have <2% success
• Network in parallel: ATS alone has 2-5% conversion, networking + ATS = 10-15%
• Follow up: Email hiring manager after applying (reference application ID)

Ethical considerations:

• ✅ Auto-fill forms with accurate info
• ✅ Generate honest, tailored cover letters
• ✅ Only apply to roles you're qualified for
• ❌ Don't lie about experience or skills
• ❌ Don't mass-apply to every posting
• ❌ Don't bypass "no bots" ToS without permission

---

How to Run a Complete Job Search

Prerequisites

1. User provides:

   • Target role (e.g., "Senior ML Engineer")
   • Industries/companies (e.g., "AI startups, series A-C, 50-200 employees")
   • Location (e.g., "San Francisco Bay Area, NYC, remote")
   • Dealbreakers (e.g., "Must have GPU budget, remote-first culture")

2. Budget:

   • Minimum: $20 SerenBucks (covers 1-2 comprehensive searches)
   • Recommended: $25 SerenBucks (covers 3-5 searches with multiple iterations)

Full Workflow Example

User says: "Help me find Senior ML Engineer roles at AI startups in San Francisco"

Phase 1: Company Discovery

python3 scripts/agent.py discover \
  --role "Senior ML Engineer" \
  --industry "Artificial Intelligence" \
  --location "San Francisco" \
  --employee_range "50-200" \
  --funding_stage "Series A,Series B,Series C" \
  --limit 50

Phase 2: Research Top 20

python3 scripts/agent.py research \
  --companies companies.json \
  --limit 20 \
  --output research.json

Phase 3: Find Hiring Managers (Apollo)

python3 scripts/agent.py find-contacts \
  --companies research.json \
  --titles "Engineering Manager,VP Engineering,Director of Engineering" \
  --tool apollo \
  --output contacts.json

Phase 3b: Enrich with Social Context (Playwright)

python3 scripts/agent.py enrich-contacts \
  --contacts contacts.json \
  --tool playwright \
  --limit 10 \
  --output contacts_enriched.json

Phase 4: Discover Events

python3 scripts/agent.py discover-events \
  --location "San Francisco" \
  --industry "AI,Machine Learning" \
  --date_range "2026-03-01,2026-04-30" \
  --output events.json

Phase 5a: Verify Emails

python3 scripts/agent.py verify-emails \
  --contacts contacts_enriched.json \
  --output contacts_verified.json

Phase 5b: Generate Outreach

python3 scripts/agent.py generate-outreach \
  --contacts contacts_verified.json \
  --background user_background.txt \
  --events events.json \
  --limit 3 \
  --output outreach.json

Phase 6: Track Applications

python3 scripts/agent.py track \
  --outreach outreach.json \
  --database applications.db

Phase 7: Automated Job Applications

python3 scripts/agent.py auto-apply \
  --companies research.json \
  --role "Senior ML Engineer" \
  --resume resume.pdf \
  --background user_background.txt \
  --limit 12 \
  --output applications.json

---

Cost Breakdown

Networking-Only Strategy (50 companies → 10 targets → 3 outreach)

| Phase | Tool | Cost | |-------|------|------| | 1. Company Discovery | AlphaGrowth | $1.50 (50 companies × $0.03) | | 2. Company Research | Perplexity + Exa | $4.40 (20 companies × $0.22) | | 3. Hiring Manager Discovery | Apollo | $4.00 (10 companies × $0.04 × 10 contacts) | | 3b. Social Context Enrichment | Playwright | $0.40 (10 top targets × $0.04) | | 4. Event Discovery | Exa + Playwright | $0.40 (10 events × $0.04) | | 5a. Email Verification | AlphaGrowth | $0.50 (50 emails × $0.01) | | 5b. Outreach Generation | Seren Models (GPT-5.2) | $9.00 (3 emails × $3.00) | | Total (Networking Only) | | $20.20 |

Full Strategy (Networking + ATS Applications)

| Phase | Tool | Cost | |-------|------|------| | 1-5b. Networking (from above) | Multiple | $20.20 | | 7. Automated Job Applications | Playwright + GPT-5.2 | $36.00 (12 applications × $3.00) | | Total (Networking + ATS) | | $56.20 |

Strategy Comparison

| Approach | Cost | Volume | Conversion | Expected Interviews | |----------|------|--------|------------|---------------------| | Networking only | $20.20 | 3 outreach | 7-40% | 0.2-1.2 | | ATS only | $36.00 | 12 applications | 2-5% | 0.24-0.6 | | Combined | $56.20 | 3 outreach + 12 apps | Blended | 0.44-1.8 |

Recommendation: Use combined strategy for maximum coverage. Networking gets you high-quality conversations, ATS gets you volume/backup.

Optimized for budget:

• Research 10 companies instead of 20: -$2.20
• Generate 2 outreach emails instead of 3: -$3.00
• Skip Playwright social enrichment: -$0.40
• Budget total: $14.60

Minimum viable:

• Research 5 companies: $1.10
• Find contacts for 3 companies: $1.20
• Generate 1 outreach email: $3.00
• Minimum total: $5.30

---

Implementation Status

✅ Fully Implemented & Working

Phase 1: Company Discovery

• ✅ AlphaGrowth company search with filters
• ✅ Export to JSON for next phase

Phase 2: Company Research

• ✅ Perplexity research integration
• ✅ Exa semantic search for technical content
• ✅ Hiring signal detection

Phase 3: Hiring Manager Discovery

• ✅ Apollo.io people search (verified emails)
• ✅ Playwright LinkedIn enrichment (social context)
• ✅ Mutual connection detection

Phase 4: Event Discovery

• ✅ Exa event search
• ✅ Playwright scraping (Eventbrite, Meetup, Luma)
• ✅ Speaker/attendee matching

Phase 5: Email Verification & Outreach

• ✅ AlphaGrowth email verification
• ✅ GPT-5.2 personalized email generation
• ✅ Template system with user background injection

Phase 6: Application Tracking

• ✅ SQLite database schema
• ✅ CRM-style tracking (outreach, responses, interviews)
• ✅ Follow-up scheduling

Phase 7: Automated Job Applications

• ✅ Playwright job posting scraper (Greenhouse, Lever, Workday, Ashby)
• ✅ GPT-5.2 resume tailoring per role
• ✅ GPT-5.2 cover letter generation
• ✅ Form automation (personal info, work history, education)
• ✅ Document upload (PDF resume + cover letter)
• ✅ CAPTCHA solving (2Captcha integration)
• ✅ Application submission and confirmation tracking

⚠️ Limitations

Manual steps:

• User must manually send networking outreach emails (no auto-send)
• User must manually update tracker with responses
• User must manually schedule interviews

Not implemented:

• Email sending via Gmail/Outlook API (security/privacy reasons)
• Automatic response parsing (requires email access)
• Calendar integration for interview scheduling
• Video interview question answering (requires human interaction)

---

Control Commands

Show Pipeline Status

python3 scripts/agent.py status --database applications.db

Output:

📊 Job Search Pipeline Status

Phase 1: Company Discovery
  • Discovered: 50 companies
  • Researched: 20 companies
  • Targeted: 10 companies

Phase 3: Hiring Manager Discovery
  • Contacts found: 47
  • Emails verified: 43 deliverable
  • Social context: 10 enriched

Phase 5: Networking Outreach
  • Emails generated: 3
  • Sent: 3
  • Responses: 1 (33.3%)
  • Interviews: 0

Phase 7: ATS Applications
  • Job postings found: 12
  • Applications submitted: 12 (100%)
  • Confirmations received: 12
  • Responses: 0
  • Interviews: 0

Phase 6: Overall Tracking
  • Total touchpoints: 15 (3 outreach + 12 applications)
  • Pending follow-ups: 12
  • Active conversations: 1
  • Interview conversion: 0% (0/15)

Update Tracker (After Sending Email)

python3 scripts/agent.py update-tracker \
  --id 42 \
  --status sent \
  --sent_date 2026-03-10 \
  --notes "Sent via Gmail, personalized with MLSys reference"

Record Response

python3 scripts/agent.py record-response \
  --id 42 \
  --responded yes \
  --response_date 2026-03-12 \
  --next_step "Coffee chat scheduled for 3/18" \
  --status active

Schedule Follow-Up

python3 scripts/agent.py schedule-followup \
  --id 42 \
  --followup_date 2026-03-17 \
  --template "polite_nudge" \
  --notes "Follow up if no response by end of week"

---

Monitoring & Logs

applications.db (SQLite)

Schema:

CREATE TABLE applications (
  id INTEGER PRIMARY KEY,
  company_name TEXT,
  company_domain TEXT,
  contact_name TEXT,
  contact_title TEXT,
  contact_email TEXT,
  contact_linkedin TEXT,
  outreach_email TEXT,
  sent_date DATE,
  response_date DATE,
  responded BOOLEAN,
  interview_scheduled BOOLEAN,
  interview_date DATE,
  status TEXT, -- pending, sent, responded, interview, offer, rejected
  followup_date DATE,
  notes TEXT,
  created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
);

outreach.jsonl (JSONL Log)

One line per generated email:

{"timestamp": "2026-03-10T14:00:00Z", "company": "Anthropic", "contact": "Sarah Chen", "subject": "Your MLSys talk + coffee at AI Safety Summit?", "body": "...", "personalization": {"hook": "MLSys talk", "value": "distributed training experience", "ask": "coffee at conference"}, "cost": 3.00}

research.json (Company Research)

{
  "companies": [
    {
      "name": "Anthropic",
      "domain": "anthropic.com",
      "culture": "Research-focused, safety-first",
      "tech_stack": ["Rust", "Python", "PyTorch"],
      "recent_news": "Launched Claude 4.6",
      "hiring_signals": ["3 job postings this week", "15% team growth"],
      "research_date": "2026-03-10",
      "cost": 0.22
    }
  ]
}

---

Best Practices

For Job Seekers

1. Quality over quantity: 3 perfect emails > 30 generic blasts
2. Research deeply: Read their blog, watch their talks, understand their tech
3. Lead with value: What can you contribute, not what you need
4. Event-based networking: In-person > cold email (40% vs 7% response rate)
5. Follow up: 70% of successful hires required 2+ touchpoints
6. Track everything: CRM discipline separates amateurs from pros

For Developers (Claude)

1. Validate email deliverability: Always run AlphaGrowth verification before outreach
2. Personalization is key: Generic emails have <5% response rate
3. Never spam: 3-5 thoughtful emails > 50 mass blasts
4. Cost transparency: Show user exact costs before running each phase
5. Privacy first: Delete contact data after job search complete
6. Follow ethical guidelines: Respect platform ToS, no scraping private data

---

Troubleshooting

"No companies found"

• Loosen filters (employee range, funding stage)
• Try different industries or locations
• Check AlphaGrowth coverage for target market

"No hiring managers found"

• Try broader title search ("Manager", "Director", "VP")
• Use Playwright as fallback to scrape LinkedIn company page
• Some companies don't list full org charts publicly

"Low response rate (<10%)"

• Improve personalization (reference specific work, posts)
• Try event-based outreach instead of cold emails
• Check email deliverability (verify sender domain, SPF/DKIM)
• A/B test different subject lines

"Email verification failed"

• AlphaGrowth may not have coverage for all domains
• Try alternate email formats (first.last@, flast@, etc.)
• Use Playwright to find email from LinkedIn profile

---

Strategy: Taariq's Approach (From Blog Post)

❌ What Doesn't Work

• Submitting resume to ATS (Applicant Tracking System)
• Applying to 100+ jobs via job boards
• Waiting for recruiters to respond
• Generic LinkedIn InMails
• Mass cold emails

✅ What Works

1. Identify decision-makers (hiring managers, not HR)
2. Network in person (conferences, meetups, office events)
3. Demonstrate value first (reference their work, show expertise)
4. Build relationships (coffee chats, not transactional asks)
5. Leverage warm intros (mutual connections > cold outreach)

Key Insight

"The hiring manager already knows they need to hire. They're just waiting to meet someone great. Be that person they meet at the conference, not resume #247 in their inbox."

---

AgentSkills.io Standard

This skill follows the AgentSkills.io open standard for agent skills, ensuring compatibility across:

• Claude Code
• OpenAI Codex
• Google Gemini
• Any compatible LLM tool

Taariq Lewis, SerenAI, Paloma, and Volume at https://serendb.com Email: [email protected]