SigOpt Summit 2021

Nick Payton
Active Search, Advanced Optimization Techniques, AI at Scale, Artificial Intelligence, BERT, Biology, Classification, Clustering, CNN, Convolutional Neural Networks, Deep Learning, Experiment Management, Finance, Fraud Detection, Gradient Boosting, Graph Neural Networks, Healthcare, Human Activity Recognition, Hyperparameter Optimization, Intelligent Experimentation, Knowledge Graphs, LSTM, Machine Learning, Materials Science, Multimetric Optimization, Natural Language, Prediction, Recommendation System, Regression, RNN, Robotics, Segmentation, Simulations & Backtests, Speech, Supervised, Time Series, Topic Modeling, Transformers, Vision

SigOpt is hosting our first user conference, the SigOpt AI & HPC Summit, on Tuesday, November 16, 2021. It is virtual and free to attend, so sign up today at Attendees will be able to join the talks and panels, meet with speakers in breakout rooms for deeper discussions and network with each other. To give you a sense of the Summit, we will publish a series of blog posts in advance of the event. This post kicks off that series, providing an overview of the event. Future posts will focus on themes that cut across the talks and panels. 

We decided to host the SigOpt Summit to showcase the great work of our customers. Our customers solve a wide variety of modeling problems and are deeply expert in these spaces. We think they have a lot to share and that many modelers could benefit from the lessons they’ve learned along the way. We also believe that many modelers could benefit from exchanging ideas with our customers and vice versa. So we are hosting the Summit to serve as a platform for showcasing this great work and are grateful to do so. 

The SigOpt Summit will include a series of individual 30-min talks broken up with hour-long panels. Although the event is virtual, you will be able to engage with the speakers in live discussion or breakout rooms after their talks. 

So who are these customers and what will they discuss? Generally, each speaker will focus on modeling problems they have addressed with an emphasis on how experimentation was critical to this process. In particular, the Summit will focus on how they implemented best practice techniques and tools that enabled them to take an intelligent approach to their experimentation and therefore solve modeling problems more efficiently at greater scale and with better results. Here is a preview of some of the topics that we will cover during the event: 

  • Pablo Zegers from Anastasia AI will discuss time series forecasting and how developing a platform to enable it has allowed them to democratize access to time series predictions for mid-market companies
  • Bhanu from Mindtree will share how they optimizing transformers from the HuggingFace library with SigOpt for natural language processing tasks that accelerated inference by over 50% and drove higher accuracy for summation
  • Subutai Ahmad from Numenta will discuss the application of sparsity and other neuroscientific techniques to evolving a novel architecture for ResNet on computer vision tasks to drive greater efficiency and robustness for performance
  • Shayan Mortazavi from Accenture will share how his team efficiently trained and deployed an ensemble of models to solve a predictive maintenance problem for their customer in the oil and gas space
  • Alexander Rosenberg Johansen from Stanford and Rafa Gomez-Bombarelli will share different biological prediction problems they solved with deep learning and how hyperparameter optimization was critical to their workflow
  • Venkatesh Ramanathan from PayPal will join Da Zheng from Amazon and Sasi Avancha from Intel Labs to discuss techniques for training and optimizing graph neural networks across massively parallel compute
  • Ke Ding and Dave Austin from Intel will share how SigOpt contributes to faster training convergence for recommender systems – and DLRM in particular – and how applying intelligent experimentation can drive better performance across a wide variety of tasks
  • Marat Latypov from the University of Arizona, Vishwanath Hegadekatte from Novelis and Paul Leu from the University of Pittsburgh will discuss how experimental design is critical driving breakthrough outcomes in their research related to physical processes
  • And Scott Clark, SigOpt CEO and Founder, will discuss tools, techniques and decisions that are the critical ingredients that go into the intelligent experimentation recipe with practical ways you can implement them in your own workflow

If you want to get a better sense of how SigOpt could impact your workflow than simply reading about use cases, sign up in seconds at If you want to learn from our customers, sign up for the SigOpt Summit for free at Look for future posts that focus more deeply on the themes that will cut across the talks and panels at the Summit. We look forward to seeing you there!

Nick Payton
Nick Payton Head of Marketing & Partnerships

Want more content from SigOpt? Sign up now.