MACHINE LEARNING,

ARTIFICIAL INTELLIGENCE &

DATA SCIENCE CONSULTING

Need A.I.? We’re ready to jump into your project!

  • Capitalize on BERKELEY’S TOP AI/ML RESEARCH

  • Bring the NEXT GENERATION’S AI/ML LEADERS to your team

  • GREAT RATES for expert-level AI/ML solutions

null

Market Analysis

Improve your marketing efficiency and business profit by analyzing consumer decisions, optimizing social media presence, and uncovering target markets.
null

Business Intelligence

Reduce operational inefficiencies by streamlining business processes and optimizing supply chain management with powerful and interpretable models.
null

Exploratory Data Analysis

Turn your unused data into profitable insights with tailored unsupervised algorithms by discovering and leveraging hidden patterns and correlations. 
null

Deep Learning

Stay ahead of the competition by processing millions of text documents and detecting objects in your photos and videos using deep learning algorithms.
null

Anomaly Detection

Detect and prevent fraud and foul play before it happens by staying ahead of anomalies in consumer patterns, financial data, or internet traffic.
null

Recommendation systems

Simplify and enhance your customers’ experience by building intelligent customer profiles and making actionable service and product recommendations.

Who We Are

We are three machine learning engineers from UC Berkeley, specializing in computer vision, language understanding, and reinforcement learning. We have published our work in peer-reviewed journals and academic conferences, won international research awards, created patentable deep learning processes for industry leaders, led educational initiatives, and worked in some of the world’s top research labs.

  • null

    Riley Edmunds

Riley Edmunds

Riley is the Head of Research at Machine Learning at Berkeley and a researcher in BAIR under Professor Dawn Song studying similarity metrics for neural network models. He has consulted in ML/AI for 3+ years.

While at the International Computer Science Institute, Riley studied the feasibility of implementing neural networks in Fourier domain. Riley handled deep learning projects at Intuit, where he authored a paper, and filed a patent over Summer of 2017. Following this, he co-authored a textbook chapter in adversarial machine learning, an ICRA conference paper on deep reinforcement learning, and also led a team to win an award for formulating adversarial attacks against meta-learning models at the 2017 Deep Learning and Security Workshop held in Singapore.

As Head of Research for ML@B, he guides research projects to completion and publication in major conferences, while also spearheading outreach and relations with academic partners at ICSI and BAIR.

Riley was born in Chicago but grew up in Lucca, Italy. In his spare time, he enjoys to read, cook and exercise. He is also training for his first triathlon this summer.

Riley is based in Berkeley, CA.

  • null

    Phillip Kuznetsov

Phillip Kuznetsov

Phillip is the founder of Machine Learning at Berkeley, an organization on the UC Berkeley campus devoted to research, delivering quality educational content, and leading consulting projects with industry partners. He is also a computer vision researcher in BAIR under Professor Alyosha Efros.

As President, Phillip leads over 75 of Berkeley’s top AI/ML minds in weekly workshops on topics in deep learning ranging from TensorFlow fundamentals all the way towards theory of Generative Adversarial Networks (GANs) and more. Through the organization, he has started numerous educational initiatives, including the Machine Learning DeCal (a student-taught course on the fundamentals of data science and machine learning).

Phillip created deep learning and computer vision models for Adobe Research over the summer of 2017. Phillip is passionate about adversarial machine learning, rendering art with deep learning, few-show learning scenarios, and neural network architecture search.

Phillip is originally from Salt Lake City, Utah. In his free time, he enjoys to rock climb and ski.

Phillip is based in Berkeley, CA.

  • null

    Noah Golmant

Noah Golmant

Noah is a consultant and a researcher at UC Berkeley primarily interested in the mathematics behind machine learning. He has worked on research and industry projects in areas including deep learning, computer vision, large-scale/distributed machine learning, natural language processing, and reinforcement learning. He has consulted in ML/AI for 4+ years.

While at Berkeley he has been heavily involved in Machine Learning at Berkeley and the RISE, AMPLab, and ASPIRE research labs. He is passionate about the theory of optimization and methods for making machine learning faster at scale. Noah studies deep learning theory and computer vision in RISELab under Professor Joseph E. Gonzalez.

Noah’s writes a blog series on understanding the mathematics underlying optimization in machine learning. Previously he helped teach CS189, Berkeley’s largest undergraduate course in machine learning, and the Machine Learning DeCal.

Noah is originally from Stafford, Virginia. In his spare time, he enjoys to read, hike, and spend quality time in nature.

Noah is based in Berkeley, CA.

Research: Awards & Patents

  • Awarded $100,000 grant from Amazon's Alexa Prize Competition

    Selected to compete as one of twelve sponsored teams (and the only undergraduate team sponsored) in the inaugural Amazon Alexa Prize competition. Built generative conversational AI model using neural turing machines and inverse reinforcement learning. Phillip Kuznetsov, James Bartlett, William Guss, Piyush Patil, 2016.

  • Filed Patent Application in Semi-Supervised Deep Learning

    Developed novel anomaly detection algorithm behind Intuit’s patent application: “Semi-supervised Clickstream Embeddings for Anomaly Detection.” while working with Intuit’s Innovation & Advanced Technology Team (I.A.T.). Riley F. Edmunds, Efraim Feinstein, 2017.

  • Awarded International Research Award at NUS, Singapore.

    For work at the intersection of Adversarial Machine Learning and Meta Learning.
    Riley F. Edmunds, Noah Golmant, Vinay Ramasesh, Philip Kuznetsov, Piyush Patil, Raul Puri at the Deep Learning and Security Workshop 2017 (DLSW ‘17) at NUS, Singapore, 2017.

Papers & Publications

Tensegrity Robot Locomotion under Limited Sensory Inputs Via Deep Reinforcement Learning.

Jianlan Luo, Riley F. Edmunds, Franklin Rice, Alice M. Agogino. Accepted at IEEE International Conference on Robotics and Automation (ICRA), 2018.

Shift: A Zero FLOP, Zero Parameter Alternative to Spatial Convolutions.

Bichen Wu, Alvin Wan, Xiangyu Yue, Peter Jin, Sicheng Zhao, Noah Golmant, Amir Gholaminejad, Joseph Gonzalez, Kurt Keutzer. Accepted as a conference paper at the Conference on Computer Vision and Pattern Recognition (CVPR), 2018.

Transferability of Adversarial Attacks in Model-Agnostic Meta-Learning.

Riley F. Edmunds, Noah Golmant, Vinay Ramasesh, Phillip Kuznetsov, Piyush Patil, Raul Puri. Accepted at Deep Learning and Security Workshop (DLSW), 2017.

Adversarial Machine Learning.

Phillip Kuznetsov, Riley F. Edmunds, Ted Xiao, Humza Iqbal, Raul Puri, Noah Golmant. Accepted as a chapter in the AI Safety and Security Textbook, 2017.

Speaking & Workshops

Workshop on “Introduction to Deep Learning.
Machine Learning at Berkeley, January, 2018.

Talk on “Transferability of Adversarial Attacks in Model-Agnostic Meta-Learning.
Deep Learning Security Workshop at NUS, Singapore, December 2017.

Workshop on “Introduction to Generative Adversarial Networks.
Machine Learning at Berkeley, November, 2017.

Workshop on “Making Art with Deep Learning.
Machine Learning at Berkeley, October, 2017.

Talk on “Deep Hierarchical Embeddings for Dynamic Targeted Anomaly Detection.
Intuit I.A.T. Lab, August 2017.

Talk on “Tunable Efficient Unitary Neural Networks.
International Computer Science Institute, July 2017.

Talk on “Complex-Valued Neural Networks: Activation Functions.
Machine Learning at Berkeley, May 2017.

Talk on “Simons Institute: Representation Learning.
Intuit IAT Lab, July 2017.

Talk on “Complex-Valued Deep Neural Networks.
Machine Learning at Berkeley, January 2017.

Talk on “Batch Methods for Incremental Learning.
UC Berkeley RISE Lab, November 2016.

Talk on “Sound Classification with Complex-Valued Neural Networks.
Machine Learning at Berkeley, November 2016.

Talk on “Environmental Sound Classification with Convolutional Neural Networks. International Computer Science Institute, September 2016.

How can      help you?  

Schedule a free 30-MINUTE CONSULTATION CALL to discuss your project ideas and business needs with an Alinea engineer. You can also reach us at contact@alinea.ai.