AutoMM
An agent-driven workflow to autonomously solve mathematical modeling problems.
An agent-driven research harness for automated math modeling and version control.
AutoMM is an agent-driven research harness built on the dsh ecosystem to autonomously execute mathematical modeling tasks. It suits researchers and students tackling structured math problems who want to automate everything from literature review to formula derivation and final reporting.
What it is
The tool drives a step-by-step math modeling workflow covering problem understanding, literature research, hypothesis formulation, calculation, and feasibility checks. It runs isolated tasks via a supervised worker framework and compiles results into a traceable final summary report while offering a real-time monitoring dashboard.
Who it's for
It suits quantitative researchers, students, and modeling competitors looking to offload repetitive pipeline execution and version control friction. It does not suit casual users looking for a plug-and-play chat interface or automated academic paper writing, as it explicitly stops short of writing full papers.
Key features
Agent-Driven Workflow
Progresses sequentially through problem understanding, formulas, implementation, and visualization.
Real-Time Dashboard
Starts a live dashboard to monitor modeling status, hypotheses, and formula derivations.
Isolated Task Execution
Creates isolated tasks handled by asynchronous supervised workers to prevent execution blocks.
Traceable Reports
Generates a structured final summary report compiling the entire pipeline output.
Up and running in 3 steps
Our honest verdict
What's great
- Automates the full pipeline from problem understanding to calculation and feasibility checks
- Includes a live dashboard for tracking modeling state and formula derivations in real-time
- Uses isolated task creation and asynchronous workers for controlled execution
- Alleviates painful version control issues inherent in multi-step modeling projects
Where it falls short
- Does not automatically write final papers, requiring manual composition steps
- Requires manual placement of real problem statements and data files before running
- License is not stated in the repository metadata
- Demands familiarity with configuration files and local script execution
Pricing & license
The software is provided via its source repository with no license stated. The only cost is running it yourself on your own hardware using your own compute resources.
Open remote AI-training roles
767 open remote AI-training roles, updated November 14, 2026; most pay $20–$175 an hour. Zekai lists them free; the application goes to the employer’s platform.
Alternatives
These reviewed AI tools do the same job instantly — from our scored directory.
Use it, or have us set it up
FAQ
Does AutoMM write the final competition paper for you?
No, the repository states that this version does not automatically write papers.
How are calculations handled during a run?
Tasks are isolated and run asynchronously via supervised workers to keep implementation agents from waiting.
Where do I put the input problem?
You must place the real problem text into request/problem.md and datasets into the data/ directory.
