Zero Cost CRM

Salesforce for teams that still live in Sheets — self-hosted pipeline, contacts, follow-ups, SDR activity, and optional call recordings. MIT. One command to run.

We don't need Salesforce. We need to know which SDR is converting and why.

Most early-stage B2B teams hire their first SDR and manage them on Google Sheets, scattered call recordings, and founder intuition. We built the system we wished existed.

Run locally · Good first issues · Fun issues · Architecture · API · Contributing · Discussions · Security · Book a demo

New here? make setup && make dev → open localhost:5173 → pick an unassigned good first issue → comment I'd like to take this. We review fast and mentor first-timers. Rewards: see #23.

---

Why this exists

We're a small team. Paying enterprise CRM prices for three reps is absurd.

We needed:

1. A 13-stage pipeline we actually use
2. Visibility into who is working (logins, active time, outcomes)
3. Follow-up discipline — connected calls without a next step light up as alerts
4. Call recordings attached to contacts, not buried in Drive folders
5. Something a founder can stand up in one command

This repo is Zero Cost CRM: pipeline operations for founder-led sales teams.

ConvoBrains is the intelligence layer: conversation analysis for call quality, pitch effectiveness, and objection handling.

Zero Cost CRM tells you what happened.

ConvoBrains tells you why it happened.

---

Run it in 3 steps

Requires Docker and Node.js 22+.

git clone https://github.com/ConvoBrains/zero-cost-crm.git
cd zero-cost-crm
make setup && make dev

Open localhost:5173 and sign in:

Email:    founder.seed@convobrains.com
Password: TestSeed123!

This is a demo seed account with 3 days of sample SDR activity. It never touches production.

Want conversation intelligence on your own calls? Book 15 min with us.

make setup installs dependencies, starts an isolated Postgres container, applies the schema, and loads companies, contacts, and three days of SDR activity.

make reset-demo   # wipe & reseed local demo data

Environment

cp .env.example .env.local          # production-style local API
cp testing/functional/.env.testing.example testing/functional/.env.testing  # demo path (make setup does this)

Important variables (see .env.example):

Variable	Purpose
DATABASE_URL	Postgres connection string
JWT_SECRET	Required — long random secret
ALLOWED_EMAIL_DOMAIN	Login allow-list (convobrains.com, comma list, or * for any)
CORS_ORIGINS	Allowed browser origins (comma-separated)
AWS_*	Optional — required only for call-recording uploads

---

Is this for you?

Use this if you:

• Have 1–10 SDRs and no Salesforce admin
• Still live in Sheets / Notion / WhatsApp for deal tracking
• Care more about daily activity + call discipline than forecasting modules
• Want self-hosted data and an open codebase

Skip it if you need enterprise CPQ, multi-currency ERP integrations, or a full marketing automation suite.

---

What's in Zero Cost CRM

Capability	What you get
Live dashboard	Follow-ups due, demos, active opps, won/lost
13-stage pipeline	Drag deals from lead → closed
Contacts	Champions, statuses, notes, LinkedIn
Paste import	Excel / Sheets / CSV → deduped companies + contacts
Call recordings	Upload & play audio per contact (S3 optional)
SDR activity	Logins, active/idle time, outcomes, targets
Manager alerts	No login by 10:30, zero connects, missing follow-ups
Roles	Founder / admin / SDR
Self-hosted	Your Postgres, your rules
Instance settings	Brand name, stages, statuses in DB (Settings UI)

Configuring your instance

Keep the code generic. Put company-specific values in env or the database:

What	Where
Postgres / JWT / email domain / S3 / CORS	Server .env (DATABASE_URL, JWT_SECRET, ALLOWED_EMAIL_DOMAIN, …)
Brand name, logo, pipeline stages, contact statuses	DB app_settings — edit in Settings (founder/admin) or PATCH /api/settings
Champion status → pipeline stage map (championStatusToStage)	API / SQL only today — PATCH /api/settings or sql/examples/convobrains-settings.sql
Discovery questions on the company form (discoveryQuestions)	API / SQL only today — same as above (not editable in the Settings UI yet)
Optional first-boot brand	BRAND_NAME, BRAND_TAGLINE, BRAND_LOGO_URL in env (seeded once)

Settings UI covers branding + stages + contact statuses. The champion sync map and discovery questions still go through the API (or the SQL example). See docs/API.md for a PATCH example.

---

The SDR operating model

Daily rhythm

1. Morning brief — open the dashboard; clear follow-ups due today
2. Pipeline — move cards only when the stage actually changed
3. Contacts — update status after every dial
4. Record — attach the call when it mattered
5. Activity (managers) — check who worked, who coasted, who needs coaching

The 13-stage pipeline

Lead Added
  → Discovery Call Done
  → Follow-up
  → Demo Scheduled
  → Demo Delivered
  → Commercial Proposal Shared
  → POC Kickoff
  → Client Data Received
  → POC Delivered
  → Final Negotiation
  → Closed Won | Closed Lost | Not Interested

Targets (editable in Activity)

Metric	Default
Calls made	80
Follow-ups set	25
Demos scheduled	4

---

Stack

React 19 + TypeScript + Tailwind
              ↓
        Express 5 API
              ↓
          PostgreSQL 16
              ↓
     S3 (optional call recordings)

AWS is only required when testing uploads. Activity, pipeline, and import work without it.

Schema is applied from sql/schema.sql (idempotent). See docs/ARCHITECTURE.md.

---

Commands

Command	What it does
make setup	Install, provision, migrate, seed
make dev	Web + API against the local demo DB
make reset-demo	Rebuild demo fixtures
make lint	Static checks
make build	Production build
make test	Unit tests
make test-api	API functional tests (DB prepared)
make help	All targets

---

Import leads

1. Open Import Leads
2. Paste from Excel, Sheets, or CSV — or upload .csv / .xlsx
3. Columns: Company · Prospect Name · Job Title · Email · Phone · Location · Employees · Industry
4. Import — companies create/update; duplicate emails skip

Samples in the UI use synthetic @*.example data only.

---

Deploy

Vercel + PostgreSQL

1. Import the repo in Vercel.
2. Set DATABASE_URL, JWT_SECRET, ALLOWED_EMAIL_DOMAIN, and optional AWS / CORS_ORIGINS.
3. DATABASE_URL='postgresql://…' npm run db:migrate
4. Deploy and hit /api/health.

Docker / VPS

# create .env with production secrets (never commit it)
make docker-build
make docker-up
make health

Root docker-compose.yml runs the app container; Postgres is external (managed DB or your own). Local demo Postgres is provided by testing/functional/docker-compose.yml via make setup.

---

Security

• Do not ship demo credentials or testing/functional/.env.testing to production.
• Use a strong unique JWT_SECRET, verified TLS for Postgres, configured CORS_ORIGINS, and private object storage.
• Report vulnerabilities privately — see SECURITY.md.

If you ever committed real secrets to git: rotate them immediately. History rewrite does not protect prior clones or forks.

---

Branding & trademarks

MIT covers the code. The ConvoBrains name, logo, and marketing assets in public/ remain ConvoBrains trademarks. Forks may keep attribution or replace branding; do not imply official endorsement.

Google Fonts (DM Sans, Instrument Serif) are loaded from Google’s CDN under their respective OFL licenses. See NOTICE.

---

Contributing

We welcome first-time PRs. Maintainers aim to reply within a day on claimed issues.

Start here	Link
Beginner tickets	good first issue (unassigned)
Showcase / delightful work	fun label
How to claim + ship	CONTRIBUTING.md
Rewards + onboarding	#23 Start here
Ideas & questions	Discussions
Hall of fame	CONTRIBUTORS.md

make setup && make dev
npm test

---

License

MIT © 2026 ConvoBrains

---

Built by ConvoBrains
Turn conversations into intelligence.

Book a demo · support@convobrains.com · LinkedIn