Skip to content

aiassistsecure/journey

BranchesTags

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
.env.example
.env.example
 
 
.gitignore
.gitignore
 
 
LICENSE
LICENSE
 
 
README.md
README.md
 
 
journey.py
journey.py
 
 
requirements.txt
requirements.txt
 
 

Repository files navigation

Journey — Journalist & Publication Finder Agent

Find the right journalists, at the right publications, with the right contact info — automatically.

An AI-powered Python agent that discovers publications covering your topic, identifies the journalists and editors who write about it, extracts their contact information (email, Twitter/X, LinkedIn), and saves everything to clean, deduplicated CSV files.

Built for founders, PR teams, and anyone who wants to get published. Point it at a topic, and it does the research for you.

Journey

Journey is an AI agent that turns what you're building into targeted media opportunities.

It finds journalists already writing about your space — so you can pitch with relevance, not guesswork.

Why Journey?

Builders ship. Most never get seen.

Journey fixes distribution.

Use cases

  • SaaS launches
  • Open source releases
  • Product updates
  • Founder stories

Philosophy

Don't spray 1,000 emails.

Find the 10 people already writing about you.

What It Does

$ python journey.py "AI startups SaaS" --max-pubs 10

============================================================
Searching for publications covering: AI startups SaaS
============================================================

  Identified 12 relevant publications.

    - TechCrunch (online media)
    - The Information (online media)
    - Bloomberg (news outlet)
    - The Wall Street Journal (newspaper)
    - Forbes (magazine)
    - PitchBook News (trade publication)
    ...

  [1/10] Looking up journalists at TechCrunch...
    Found 8 journalist(s):
      - Russell Brandom (AI Editor)
      - Rebecca Bellan (Senior Reporter)
      - Julie Bort (Venture Editor)
      ...

  [2/10] Looking up journalists at The Information...
    Found 10 journalist(s):
      - Amir Efrati (Co-Executive Editor)
      - Katie Roof (Deputy Bureau Chief, VC)
      ...

============================================================
Enriching contact information...
============================================================

============================================================
PROGRESS REPORT
============================================================
  Publications:    10 total, 8 scanned, 2 pending
  Journalists:     39 total
  With email:      16
  With Twitter:    10
  With LinkedIn:   5
  Has any contact: 22
  Needs enriching: 17
============================================================

Real results from a single run — 10 publications, 39 journalists, 22 with verified contact info, in under 2 minutes.

Example Output

Publications Found

Publication Type Journalists Found
TechCrunch online media 8
The Information online media 10
Bloomberg news outlet 5
Forbes magazine 5
PitchBook News trade publication 4
Andreessen Horowitz (a16z) blog 4
The Wall Street Journal newspaper 2
AI Magazine magazine 1

Sample Contacts

Name Title Publication Contact Found
Russell Brandom AI Editor TechCrunch email
Julie Bort Venture Editor TechCrunch email, twitter
Amir Efrati Co-Executive Editor The Information email, twitter
Katie Roof Deputy Bureau Chief, VC The Information twitter
Rachel Metz AI Reporter Bloomberg email, twitter
Alex Konrad Senior Editor Forbes twitter, linkedin
Rosie Bradbury Senior Reporter PitchBook News email, linkedin
Kate Clark Reporter The Wall Street Journal twitter

Features

  • Publication Discovery — Searches the web and uses AI to identify the most relevant newspapers, magazines, online media, blogs, and trade publications covering your topic.
  • Journalist Lookup — For each publication, finds journalists, editors, reporters, and writers who cover your beat. Extracts editorial/team pages and cross-references with web search results.
  • Contact Enrichment — Searches for email addresses, Twitter/X handles, and LinkedIn profiles. Batched in groups of 5 per publication for speed.
  • Resumable Runs — Saves progress after every publication. If a run is interrupted, re-run the same command and it picks up exactly where it left off. No duplicate work, no wasted API calls.
  • Enrich-Only Mode — Already have journalists but need more contact info? Run --enrich-only to skip discovery and just fill in missing contacts.
  • Persistent CSVs — Each topic gets a single pair of CSV files that grow over time. New publications and journalists are appended; existing data is never overwritten.
  • Progress Reports — See at a glance how many publications are scanned, how many journalists are found, and how many still need contact enrichment.
  • Retry with Backoff — API calls automatically retry on transient errors (429, 5xx) with exponential backoff.
  • Smart Deduplication — Publications deduped by name + domain, journalists by normalized name + publication. Handles punctuation, casing, and whitespace.
  • Data Sanitization — All data is cleaned and normalized before saving. No nulls, no stray whitespace, no garbage in your CSV.

Quick Start

(linux, windows is slightly different venv)

git clone https://github.com/aiassistsecure/journey.git
cd journey
python3 -m venv venv
source venv/bin/activate   # different on Windows depending on which shell
pip install -r requirements.txt
cp .env.example .env        # Add your AiAssist.net API key
python3 journey.py "your topic here"

Requirements

  • Python 3.10+
  • requests library
  • An AiAssist.net API key (any plan, start free 7 days) with access to:
    • /v1/chat/completions — AI analysis (GPT-5.4 via OpenAI provider)
    • /v1/search — Web search
    • /v1/web/extract — Web page extraction

Usage

Full Scan

Discover publications, find journalists, enrich contacts — the full pipeline:

python journey.py "AI startups SaaS artificial intelligence"

Limit Publications

Focus on the top N most relevant publications:

python journey.py "AI startups SaaS" --max-pubs 10

Enrich Only

Already ran a scan but want to fill in more contacts? Skip discovery entirely:

python journey.py "AI startups SaaS" --enrich-only

Resume

Interrupted mid-run? Just run the same command again. It picks up where it left off:

# Run 1 — gets through 5 of 10 publications, then times out
python journey.py "AI startups SaaS" --max-pubs 10

# Run 2 — resumes at publication 6, skips the first 5
python journey.py "AI startups SaaS" --max-pubs 10

# Run 3 — all 10 done, just enriches any remaining contacts
python journey.py "AI startups SaaS" --max-pubs 10

Output Schema

All output is saved to the output/ directory (gitignored). Each topic gets two persistent CSV files.

Publications — output/pubs_<topic>.csv

Column Description
name Publication name
url Main website URL
type newspaper, magazine, online media, blog, trade publication
relevance Why this publication covers your topic
editorial_page_url URL of their about/team/staff page

Contacts — output/contacts_<topic>.csv

Column Description
name Journalist's full name
title Role — Senior Reporter, Editor-in-Chief, Staff Writer, etc.
publication Which publication they work at
publication_url Publication's website
beat Topics they cover
email Email address (verified from web sources)
twitter Twitter/X handle
linkedin LinkedIn profile URL
source_url Where the journalist info was found

How It Works

┌─────────────┐     ┌──────────────┐     ┌────────────────┐     ┌─────────────┐
│   Search     │────>│   Analyze    │────>│    Extract     │────>│   Enrich    │
│              │     │              │     │                │     │             │
│ Web search   │     │ GPT-5.4      │     │ Editorial      │     │ Batch web   │
│ for pubs     │     │ ranks &      │     │ pages, staff   │     │ search for  │
│ covering     │     │ identifies   │     │ directories,   │     │ email,      │
│ your topic   │     │ relevant     │     │ bylines →      │     │ Twitter,    │
│              │     │ publications │     │ journalists    │     │ LinkedIn    │
└─────────────┘     └──────────────┘     └────────────────┘     └─────────────┘
                                                                       │
                                                                       v
                                                              ┌─────────────┐
                                                              │    Save     │
                                                              │             │
                                                              │ Dedupe,     │
                                                              │ sanitize,   │
                                                              │ write CSV   │
                                                              │ (resumable) │
                                                              └─────────────┘
  1. Search — Runs 3 parallel web searches with different query angles to cast a wide net.
  2. Analyze — GPT-5.4 analyzes all search results to identify and rank the most relevant publications.
  3. Extract — For each publication, searches for journalists by name, extracts editorial/team pages, and cross-references to build a verified list.
  4. Enrich — For journalists missing contact info, runs batched web searches (5 per batch) and uses AI to extract verified contact details.
  5. Save — Results are deduplicated, sanitized, and saved to CSV after each publication. Nothing is lost if the process is interrupted.

File Structure

journey/
├── journey.py          Main entry point — the agent
├── requirements.txt    Python dependencies
├── .env.example        Template for API key setup
├── .gitignore          Excludes output/, .env, __pycache__/
├── output/             CSV output directory (gitignored)
│   ├── pubs_<topic>.csv
│   └── contacts_<topic>.csv
├── README.md
└── LICENSE             MIT

Technical Details

Detail Value
AI Model GPT-5.4 (via AiAssist.net OpenAI proxy)
Search Depth advanced (full page analysis)
Max Content Extraction 15,000 characters per page
Enrichment Batch Size 5 journalists per API call
Retry Policy 3 attempts, exponential backoff (1.5x)
Retryable Status Codes 408, 409, 429, 500, 502, 503, 504
Connection Pooling requests.Session reused across all calls
Deduplication Publications by name + domain, journalists by normalized name + publication
Data Integrity All fields sanitized, nulls converted to empty strings, whitespace stripped
Fabrication Policy Never — only includes details verified from web sources

Use Cases

  • Founders — Find journalists who cover your space to pitch your startup story.
  • PR Teams — Build targeted media lists for press releases and product launches.
  • Content Marketers — Identify publications accepting guest posts or contributed articles.
  • Researchers — Map the media landscape covering a specific industry or topic.
  • Job Seekers — Find editors and hiring managers at publications you want to write for.

License

MIT — see LICENSE for details.

About

Meet Journey! The agent that finds journalists already covering the topics you’re building, shipping, and open-sourcing.

Resources

Readme

License

View license
Activity
Custom properties

Stars

2 stars

Watchers

0 watching

Forks

0 forks
Report repository

Releases

No releases published

Packages

 
 
 

Contributors

Languages