Tools chosen for the problem.

Open-source first. Self-hosted where privacy and cost matter. Nothing on this page is here because it trends well — each of these is running in something we maintain.

Technologies
59
Categories
9

The stack, by category

Backend

9 items

  • Laravel 11
  • PHP 8.3
  • Node.js
  • Python
  • Express
  • FastAPI
  • REST
  • GraphQL
  • Horizon

Frontend

5 items

  • Blade
  • React
  • Next.js
  • Tailwind
  • TypeScript

AI / ML

8 items · self-hosted

  • Ollama
  • DeepSeek
  • LLaMA
  • Qwen
  • Whisper
  • WhisperX
  • RAG
  • Embeddings

Databases

4 items

  • MySQL
  • MongoDB
  • PostgreSQL
  • Redis

Search

3 items

  • Typesense
  • Vector search
  • Semantic search

Automation

6 items

  • n8n
  • Webhooks
  • Google Sheets
  • Zoho
  • Xero
  • Shopify Flow

DevOps

7 items

  • Ubuntu
  • Nginx
  • PM2
  • Docker
  • SSL
  • PHP-FPM
  • Cron

Marketing & APIs

5 items

  • Google Ads API
  • GA4
  • Meta API
  • Schema.org
  • JSON-LD

Voice

5 items

  • Asterisk
  • SIP
  • AMI
  • IVR
  • Diarization

Selection principle

Why so much of this is self-hosted.

Three reasons, in order of how often they decide it. Privacy: customer records, call recordings and order data stay inside infrastructure you control. Cost: per-token and per-seat pricing scales badly once a system is genuinely used. Longevity: an open-source component can still be patched in three years, whatever happens to a vendor's roadmap.

Where a hosted service is genuinely better, we say so and use it. The point is that the choice is made deliberately, per project, and written down.

See how it applies per track →

Stack review

Not sure the stack you have is the stack you need?