I Travel Often, So I Built an AI Copilot for Comparing Flights: Meet AirGem

Description and use

AirGem is an open‑source, privacy‑first flight‑search copilot that turns natural‑language requests into a concise shortlist of flights, respecting budget, timing, and calendar constraints while keeping calendar data local and transient.

Problem

Frequent travelers in Brazil need to compare flight costs, times, stops, dates, personal preferences, and calendar conflicts without opening many tabs and without giving travel platforms ongoing access to their calendars.

Outcome

AirGem enables frequent travelers to quickly obtain a concise, explainable shortlist of flights that respect budget, timing, and calendar constraints while keeping all sensitive data local and transient.

Workflow

• 1. User submits a natural‑language travel request via the web form.
• 2. Gemma parses the request into structured preferences.
• 3. SerpApi retrieves current flight search results.
• 4. Python code filters, ranks, normalizes, and checks calendar conflicts using the uploaded .ics file.
• 5. Gemma synthesizes an explanation of the chosen options.
• 6. The UI displays the shortlisted flights, alternatives, booking links, and conflict information.

Evidence

AirGem parses the request, searches current flight data, evaluates candidates, considers calendar conflicts, and explains the selected options.

Source

The calendar is used to rule out conflicts during the search, then discarded with the rest of the transient session data.

Source

Gemma parses the user's request into structured travel preferences.

Source