Our Work


A Few Recent Clients


Our team developed a real-time object detection platform using state-of-the-art deep learning to detect and track objects in live video streams. This work enabled the launch of the C&A's flagship product.

We worked with Metacog to bridge the gap between Artificial Intelligence and automating educational assessment by optimizing human learning and performance. Using this AI, the student improves their technical, college, and career readiness.

SFL worked with Cybric to add machine intelligence on top of their state-of-the-art orchestration platform. We developed a data science platform to predict vulnerabilities in various cybersecurity related scenarios.

cunesoft_logo_900px (1).png

SFL performed anomaly detection in time-series energy grid data to detect fraudulent customers in the Danish SE energy company. We used dynamic time-warping and statistical anomaly detection methods to automatically detect fraudulent customer activity.

Our team worked with Cunesoft to create a novel Information Extraction algorithm to solve their data science requirements. We built custom Natural Language Programming software that successfully extracted ten classes of pertinent information in seven different languages. 

SFL developed a set of machine learning models to determine the top attributes that are most deterministic of winning a client deal. Our work enabled the Bullhorn to optimize their sales process and maximize sales.


SFL helped WinkHealth create a machine learning algorithm to detect sleep apnea from raw audio data. A probabilistic score was provided for events in the data that were suspected apnea events. Our algorithm allowed automatic detection of sleep apnea, which can be used to replace expensive and time-consuming sleep studies performed at clinics.

SFL created an information extraction system to correctly and automatically identify contractual legal information core to their business model. 

SFL performed a market segmentation analysis for a large Japanese retailer. The dataset consisted of millions of historic purchases for tens of thousands of clients. Seven segments were created using a hierarchical clustering algorithm and a deep profile was created for each.