SigOpt is an Intel company – learn more

Sign Up
Research

Our research team is constantly developing new optimization techniques for real-world problems.

Learn More
Papers

Our team’s peer-reviewed research.

Publications

Research from our academic users.

Patents

For Academia

Use SigOpt for reproducible research for free

Why SigOpt?

SigOpt automates model optimization so your experts can focus on applying their domain expertise to your most pressing challenges.

Designed to meet your particular modeling needs

  • Earn a Return on AI & ML

The most advanced Optimization Engine that delivers better results, faster and cheaper, than all standard tuning methods

  • Tune Any Model 

Black-box optimization that efficiently optimizes any machine learning, deep learning, artificial intelligence, simulation, or other model of your choosing

  • Retain Flexibility

Our API-enabled solution easily plugs into any model management platform or model development framework so you retain flexibility in your technology stack

Learn how Two Sigma leverages SigOpt in their modeling processes

Select the right optimization approach for you

A short description of the table's data for accessibility with screen readers, etc.
 Grid SearchRandom SearchOpen Source

Models

1-4 Parameters

5-10 Parameters

11-100 Parameters

Any Model Type

Ease of Use

Low Maintenance

Reliability at scale

Organizational control

Experiment analysis

API-enabled

Priorities

Performance

Expert productivity

Compute utilization

Wall-clock efficiency

Variety of use cases

Competing objectives

Constraints

Conditional parameters

Long training cycles

Proven Results

By optimizing each model, teams who invest in automated hyperparameter optimization will amplify the performance and impact of models on their business. SigOpt’s customers also typically realize higher performance from models tuned with our solution, and, in certain cases, use SigOpt to solve challenges that unlock entirely new business opportunity.

All Resources

Accelerate Wall-Clock Time to Tune

SigOpt’s proprietary solution efficiently explores and exploits your search space to uncover the global optima for any model much faster than grid search, random search or open source Bayesian optimization.

Increase Computational Efficiency

Tune models with a computationally efficient optimization solution and maximize the utilization of your clusters. This process is critical to scaling machine learning in production.

Improve Model Performance

While there is no free lunch, SigOpt consistently outperforms other optimizers across any variety of models that your team develops to solve different business challenges.

Learn More About Our Optimization Approach

Optimization is in our DNA. SigOpt was founded by the creator of the Metric Optimization Engine (MOE) and our research team has spent years building and refining our ensemble of applied optimization algorithms to scale with our enterprise customers’ needs. Built by experts for experts, we have a deep appreciation for the broader research community from which we draw inspiration and to which we regularly contribute.

Learn More
Solutions Collage