Getting Started¶
Get your first optimization problem solved in under 5 minutes.
Prerequisites¶
- Python 3.10 or higher
- An OpenAI API key (get one here)
Quick Start¶
1. Set your API key¶
2. Choose your interface¶
The desktop app provides the best experience with a visual interface.
This downloads and runs Savanty without permanent installation.
Start a local web server with a browser-based interface.
Then open http://localhost:8000 in your browser.
Solve problems directly from your terminal.
3. Describe your problem¶
Enter your optimization problem in plain English. For example:
Text Only
I need to schedule 4 nurses (Alice, Bob, Carol, Dave) for morning,
afternoon, and night shifts over the next week.
Rules:
- Each shift must have exactly one nurse
- No one can work more than 5 shifts per week
- No back-to-back shifts for anyone
- Alice cannot work night shifts
4. Answer clarifying questions¶
If Savanty needs more information, it will ask targeted questions:
- "How many days are in the schedule?"
- "Can a nurse work the same shift type on consecutive days?"
Provide the answers, and Savanty will continue solving.
5. Get your solution¶
Savanty returns:
- Solution - The optimal assignment that satisfies all constraints
- ASP Code - The logic program used to find the solution
- Visualization - A visual representation of the result
Understanding the Results¶
Solution¶
The raw output showing which assignments were made:
Text Only
assign(alice, monday, morning)
assign(bob, monday, afternoon)
assign(carol, monday, night)
...
ASP Code¶
The Answer Set Programming code that encodes your problem:
Prolog
% Facts
nurse(alice). nurse(bob). nurse(carol). nurse(dave).
shift(morning). shift(afternoon). shift(night).
day(monday). day(tuesday). ...
% Rules
1 { assign(N, D, S) : nurse(N) } 1 :- day(D), shift(S).
% Constraints
:- assign(N, D, S1), assign(N, D, S2), S1 != S2.
Visualization¶
A formatted display of your solution, such as a schedule table or assignment chart.
Next Steps¶
- Installation Guide - Permanent installation options
- User Guide - Learn all interface options
- Examples - See more problem types
- When to Use Savanty - Understand problem suitability