Blog Alternatives
FR EN ES

The Caresse blog

Relationships, intimacy, exploration.

Articles to help you communicate better, reconnect with your partner, or simply take time for yourself. And, further down, the technical behind-the-scenes of building Caresse.

🎧

Intimacy and visual impairment: where audio fits in

Why so many intimate services assume you can see the screen, what an audio-first experience changes, and how to tell audio content apart from an accessible interface.

Read the article
📊

State of AI Romance 2026

A data report on AI, immersive audio, personalization and privacy — 20+ public sources cited, from Pew Research to Cisco and McKinsey.

Read the report
🕰️

How to Rekindle the Spark in a Long-Term Relationship

7 concrete, realistic ideas, no pressure, no miracle fix, to bring back desire and surprise in a long-term relationship.

Read the article
💬

How to Talk About Your Fantasies Without Awkwardness

Fear of judgment blocks more couples than lack of desire. A concrete method for bringing up the topic without awkwardness.

Read the article
🌙

Solo Exploration: Letting Go of Guilt

Taking time for yourself, alone, isn't selfish. Where the guilt comes from and how to let go of it.

Read the article

Building Caresse

Behind the scenes, technically.

Building Caresse #1 — Hexagonal Architecture

Hexagonal Architecture for a Multi-Provider AI Pipeline

Ports, adapters, and a real-world provider registry in .NET 10 that lets us swap LLM and TTS vendors without ever touching the domain.

Read on Medium
Building Caresse #2 — Orchestrating a Multi-Phase LLM Pipeline

Orchestrating a Multi-Phase LLM Pipeline

One user prompt, 8 sequential stages, a different model per stage, and a self-critique step that improves the output before it ever reaches TTS.

Read on Medium
Building Caresse #3 — Integration Testing a .NET API

Integration Testing a .NET API Without Touching Production

Testcontainers for PostgreSQL, Azurite for Azure Blob, a custom auth handler replacing Firebase, and WireMock.Net for RevenueCat, LLM providers and TTS — all wired into one xUnit suite that runs in CI with zero external dependencies.

Read on Medium