Brew Logging with Python

Note

Who’s this for?
Anyone who juggles coffee mugs and pip install commands—home baristas, data-curious developers, or both.

Why Track Your Brews?

Keeping a brew log turns fuzzy taste memories into searchable data. You’ll discover that the “wow” shot you pulled last Friday had 19 g in / 38 g out in 27 s at 94 °C, not just “kinda faster than usual.” Patterns appear; skill improves.

Tip

First step: Pick one variable to optimize (dose, temperature, or time). Too many knobs at once will drown your palate and your dataset.

Setting Up the Roaster’s Notebook

Create a new virtual environment.
Install two lightweight libs:
```
pip install sqlite-utils matplotlib
```
1. Spin up an empty SQLite database called brewlog.db.

Avoid bloated stacks

A single-file SQLite DB beats a heavyweight cloud service when you’re half-awake at 6 a.m.

Designing the Brew Schema

Table: brews
─────────────
id            INTEGER  (PK)
date_time     TEXT     (ISO 8601)
bean          TEXT
method        TEXT     (espresso | pourover | coldbrew …)
dose_g        REAL
yield_g       REAL
brew_time_s   INTEGER
temp_c        REAL
rating_10     INTEGER
notes         TEXT

Note

Keep notes free-form. Flavor language is messy; your schema shouldn’t police it.

A 30-Second Logging Script

from datetime import datetime
import sqlite3

def log_brew(bean, method, dose, yield_, time_s, temp, rating, notes=""):
    with sqlite3.connect("brewlog.db") as db:
        db.execute(
            """
            INSERT INTO brews
            (date_time, bean, method, dose_g, yield_g, brew_time_s, temp_c, rating_10, notes)
            VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?)
            """,
            (
                datetime.now().isoformat(timespec="seconds"),
                bean,
                method,
                dose,
                yield_,
                time_s,
                temp,
                rating,
                notes,
            ),
        )
    print("☕️ Logged:", bean, rating, "/10")

# Example
log_brew(
    bean="Ethiopia Guji",
    method="espresso",
    dose=18.5,
    yield_=37,
    time_s=29,
    temp=93.5,
    rating=9,
    notes="Blueberry jam, silky body",
)

Tip

Alias the script in your shell so that logbrew is only one keystroke away from caffeine.

Charting Your Caffeine Arc

import sqlite3, matplotlib.pyplot as plt
from datetime import datetime

with sqlite3.connect("brewlog.db") as db:
    rows = db.execute(
        "SELECT date_time, dose_g FROM brews ORDER BY date_time"
    ).fetchall()

dates = [datetime.fromisoformat(r[0]) for r in rows]
doses = [r[1] for r in rows]

plt.plot(dates, doses, marker="o")
plt.title("Daily Dose (g) Over Time")
plt.xlabel("Date")
plt.ylabel("Dose per Brew (g)")
plt.show()

Question

But where are the caffeine mg? Multiply dose by bean-specific extraction ratios if you want precision, or keep grams as a simple proxy. The trend matters more than the absolute number.

Marrying Obsidian & Python

Store the script in your Obsidian vault inside a _scripts/ folder.
After each brew, run logbrew from the vault’s root.
Embed charts via the Obsidian Image Embed or the Dataview plugin.

![[brew_chart.png]]

[!TIP] Add a Daily Note template that queries your brews table and surfaces yesterday’s extraction alongside today’s todo list.

⸻

Closing Sip

Logging brews is less about chasing perfection and more about tasting your learning curve. Each record is a breadcrumb back to a flavor memory, and every plotted line is a quiet affirmation that you’re dialing in, cup by cup, commit by commit.

Next experiment

Compare pour-over flow rates with a Bluetooth scale & integrate the data via a WebSocket. Coffee meets IoT—stay tuned!