beginner15 min5 steps

Getting Started with Squishy: Automate Sourdough Fermentation Monitoring

Learn how to use Squishy's AI vision engine to analyze sourdough starter health, predict peak fermentation times, and eliminate guesswork from your baking routine.

By AI Indigo Team

This tutorial uses:

🔍

Squishy

AI-powered sourdough starter analysis - snap a photo and know exactly when your starter is ready to bake!

Install and Configure the Squishy Mobile App

Begin by downloading the Squishy application from the App Store or Google Play. Upon launching the app for the first time, you will be prompted to create a user profile. Unlike traditional recipe apps, Squishy requires specific environmental data to calibrate its machine learning models. You must input your current ambient temperature and humidity, as these factors significantly impact yeast activity. The app uses this baseline data to adjust its fermentation prediction algorithms. Navigate to the 'Settings' tab and enable 'Push Notifications' for critical alerts, such as when your starter reaches peak volume. This ensures you don't miss the narrow window for optimal baking. If you have a smart home thermostat or humidity sensor compatible with the Squishy API, you can link these devices in the 'Integrations' menu for automated data logging, though manual entry is sufficient for beginners.

Pro Tip

Enable 'High-Res Mode' in camera settings within the app upload screen. The computer vision model relies on subtle surface textures to detect gluten structure development, which low-resolution images may obscure.

Calibrate Your Starter Profile

Before taking your first analysis photo, you need to establish a baseline for your specific sourdough starter. Each starter has a unique microbial ecosystem. In the app, select 'Create New Starter' and name it (e.g., 'Grandma's Rye'). You will be asked to input the flour type (e.g., 100% whole wheat, unbleached AP) and the hydration percentage (e.g., 100%, 75%). Squishy’s algorithms use these parameters to set expected expansion rates. For example, a 100% hydration rye starter behaves differently than a stiff whole wheat levain. Next, take a 'Day 0' photo of your starter immediately after feeding. This image serves as the control variable for future comparisons. The AI will analyze this initial state to establish a reference point for volume and surface tension. Without this calibration step, the AI cannot accurately calculate growth percentages or predict peak times relative to your specific culture’s behavior.

Pro Tip

Be precise with hydration percentages. Even a 10% variance in hydration can shift fermentation windows by hours. Use a digital scale to measure your flour and water ratios accurately before logging them.

Capture and Analyze Starter Progress

As your starter ferments, you will need to provide visual data to the AI. Place your starter jar on a flat, well-lit surface with a neutral background (white or light gray works best). Open Squishy and select your active starter profile. Tap the camera icon to snap a photo. Ensure the jar is centered and the lighting is even; avoid harsh shadows or direct sunlight, which can create glare and confuse the computer vision model. Once the photo is taken, Squishy processes the image using its on-device ML model. Within seconds, you will receive a dashboard view showing current volume expansion percentage, surface bubble density, and estimated time to peak. The app highlights areas of high activity with green overlays and lags with red. This visual feedback helps you understand not just *if* it's ready, but *how* active the culture is. Repeat this process every 30-60 minutes during active fermentation to refine the AI's prediction accuracy.

Pro Tip

Use the 'Grid Guide' feature in the app to ensure the jar remains in the same position for each photo. Consistent framing allows the AI to track volume changes more precisely by comparing pixel density across frames.

Interpret Real-Time Fermentation Alerts

Squishy moves beyond simple volume measurement by providing contextual alerts. When the AI detects that your starter has reached approximately 80% of its predicted peak volume, it will send a 'Pre-Peak' notification. This is your cue to prepare your dough mixing ingredients, as the starter will reach its maximum volume shortly. The app provides a 'Readiness Score' from 0 to 100. A score above 90 indicates peak activity, where the starter is full of gas and aromatic compounds. If you miss the peak, the AI will issue a 'Over-Fermentation' warning, showing signs of collapse or excessive acidity. Understanding these alerts is crucial for scheduling your baking. You can set the 'Baking Workflow' in the app to automatically suggest recipes that match the current state of your starter, ensuring you always have a plan when the starter is ready. This integration reduces decision fatigue and ensures you use the starter at its most potent stage.

Pro Tip

Do not ignore the 'Acidity Trend' metric. Squishy estimates pH levels based on surface color changes. If acidity is rising too fast, the AI will suggest a more frequent feeding schedule to prevent over-acidification.

Advanced: API Integration for Custom Dashboards

For developers or advanced users who want to integrate Squishy’s data into a custom smart home dashboard, the Squishy API offers robust endpoints. You can generate an API key from the 'Developer' section of your profile. Use this key to fetch real-time fermentation data via REST calls. Below is a Python example using the `requests` library to retrieve the current status of a starter identified by its unique ID. This allows you to build custom notifications or log data to external spreadsheets for long-term trend analysis. The API returns JSON data including volume expansion, estimated peak time, and confidence intervals. This is particularly useful for serious bakers who want to correlate baking results with precise environmental data over weeks or months, creating a personalized dataset that refines the AI's predictions for their specific kitchen environment.

Pro Tip

Rate limits apply to free tier API keys (10 requests per minute). Implement caching in your custom application to avoid hitting these limits during active fermentation monitoring.

Try Squishy →

🔥Stay ahead of the AI curve

Never Miss a Breakthrough AI Tool

Get the hottest AI tools, exclusive tutorials, and insider tips delivered to your inbox every Friday. Free forever.

🔒 No spam, unsubscribe anytime. We respect your inbox.

AI Tools

Free Tools

Weekly

Updates