Marketing Automation Engine

Analyzes all your Atlassian Marketplace data, generates contacts/deals, and puts them into HubSpot.

More specifically:

1. Downloads and analyzes HubSpot and Atlassian Marketplace data
2. Generates contacts from all License/Transaction contact info
3. Identifies and flags Contact-Type for each Contact/Company
4. Matches up related MPAC events via similarity-scoring
5. Updates Contacts based on match results
6. Generates Deals based on match results
7. Upserts all generated/updated HubSpot data entities

This runs in an ENV-configurable loop.

Read about the engine logic in detail in docs/ENGINE.md.

Implemented in Node.js (TypeScript) and can build a Docker image.

---

HubSpot Setup

See docs/HUBSPOT.md.

Dev Setup

1. Install Node.js 16+ and NPM 7+
2. Copy .sample.env to .env and set values.
3. npm install

Running

$ npm run once -- --in=remote --out=remote

CLI Options

--in   local | remote
    Whether to use disk-cached values, or download live data

--out  local | remote
    Whether to cache output to disk and log, or upload live data

--cached-fns  scorer.json
    (Optional) Reuse cached results of given function

--loglevel    error | warn | info | verbose | detailed
    (Optional) What the engine should log to console.log()

Developer NPM commands

# Run engine once
$ npm run once -- [options]

# Example of dry-run, using local data and cached scorer data, with medium verbosity
$ npm run once -- --in=local --out=local --cached-fns=scorer.json --loglevel=info

# Download live data and cache to disk
$ npm run download

# Run unit tests
$ npm run build # either build once
$ npm run watch # or watch and build
$ npm test

# Run engine multiple times,
#   starting with cached data,
#   and pumping output of each run
#     into input of next run
$ npm run multiple

# Explain what the engine does given certain SENs or transactions
# (Requires the engine to have been run on latest data locally)
$ npm run explain -- [--verbose] <SEN12345ABCDE>... | <transactions.json>

Running during Development

Running the engine live (steps above) will cache data locally in git-ignored data directory. After it's cached, you can use --in=local for faster development and to avoid API calls.

Instead of uploading to Hubspot, you can use --out=local and --loglevel=verbose (the default) to print data to console that would have been uploaded, or --loglevel=info to just show array counts.

After running the engine, to test logic after the Scoring Engine runs, pass --cached-fns=scorer.json to reuse the most recently results of the Scoring Engine.

Changelog

Unreleased

• Renamed npm start to npm run once (Use Docker image for continuous background processing)
• Renamed ENV variables MPAC_PASS to MPAC_API_KEY

0.1.0 (2021-11-25)

• First public version