Enterprise Data Science and ML Solutions

Let's Talk Top Developers Clutch Badge 2020
BCG Boston Consulting Group
Index Ventures
Canada Natural Resources - Canadian Forest Service
Fortune 500 company
Triple A

World-class solutions for enterprise development challenges.

Appsilon provides innovative data analyticsmachine learning, and managed services solutions for Fortune 500 companies, NGOs, and non-profit organizations. We deliver the world’s most advanced R Shiny applications, with a unique ability to rapidly develop and scale enterprise Shiny dashboards. Our proprietary machine learning frameworks allow us to deliver Computer Vision, NLP, and fraud detection prototypes in as little as one week.

Above all, we are committed to making a positive impact on the world. Through our AI For Good Initiative, we routinely contribute our skills to projects that support the preservation of human life and the conservation of animal populations all over the globe. Recently, our team has worked to mitigate poaching in Africa with computer vision, provide satellite image analysis for assessing damage after natural disasters, and build tools to help with COVID-19 risk assessment. Appsilon is also a pioneer in open source. Our packages are actively used by global organizations such as Ubisoft, Bank of America, and Renault.

Appsilon is a proud RStudio Full Service Certified Partner.

Acquisition, Validation, and Analysis
Machine Learning
Machine Learning
Computer Vision, Satellite Imagery Analysis, and NLP
Data Visualization, R Shiny Dashboards, and Python Dash

Case studies:

Optimization in port management

Maritime AIS Safety Cargo management

Working with ports drives the efficient flow of goods. This foundational business has experienced an explosion in the number of data points available thanks to developments in sensors, IoT, software, and increasing global complexity. In this visualization, you can see how we expedite the process of loading and unloading goods.

Route prediction can help both ports and crew, as we inform everyone ahead of time about incoming vessels and help captains choose the most time and cost effective ports. Traders can also benefit by having real time updates, taken from Automatic Identification System (AIS) data. 

  • Combines Satellite and AIS data in real time
  • Forecast behavior
  • Improves safety and efficiency

Automatic Valuation Models

Real Estate AVM Data Scraping

There is a famous case of a data scientist helping find a cheaper location for a new firehouse, whilst decreasing reaction times to all locations served. We do that, at scale, with the help of Automated Valuation Models. You can see differences real estate prices of specific plots and subplots in New York City. Changes in time and trends can be seen on a granular level. Expert knowledge is incorporated into the models and improves results by giving us more than data, but also years of experience.

Our approach to combining models with highly functions user interfaces drastically improves the usefulness of our models. Anyone from developers, facility managers to real estate investors can benefit from our tool. We model and predict future trends to help municipal planners, developers and even citizen investors looking for a tangible nest egg.

  • Intuitive and functional
  • Forecast future behavior
  • Augmented with Expert Knowledge

Satellites image analysis in agriculture

Image recognition Deep learning KERAS

Satellites follow Moore’s Law, in that the resolutions and image frequency exponentially increase and improve over time. We take advantage of such satellite imagery and deep learning to automatically recognize images and discover insights. The case you can see is from an agricultural plot of land where our model has identified three lost areas of corn. Farmers can use this when calculating their losses and identifying areas of improvement.

Such technology is not restricted to agriculture, but can be applied to a multum of industries. We can help shopping malls and offices extract insights from their parking lots. We can identify and track the migration of groups of land and sea animals. We are not yet at live tracking, but regular and recurring images are enough to benefit and add new business use cases to your company while staying ahead of the competition.

  • 30cm per pixel
  • Minimal travel and on-site costs
  • Deep Learning Image Recognition

Shiny Enterprise Dashboards

Shiny Data visualization Enterprise Dashboards

We build scalable & reliable custom enterprise dashboards. We focus on UX and a modern aesthetic that improves the functionality of our models while minimizing the learning curve needed for the end user. We create open source packages for Shiny that are used by thousands of data scientists around the globe.

We have experience in creating secure & dynamic dashboards in Shiny for hundreds of users.  Our custom packages introduce such features as authentication with LDAP integration, usage analytics and multilingual interface.

  • Hundreds of simultaneous users
  • 7 days to mock up
  • Data acquisition automation
BCG Boston Consulting Group
John Dannberg, Principal
The Boston Consulting Group

Appsilon Data Science proved to be an excellent business partner. The high technical skillset coupled with a solid business understanding made the cooperation flawless. Appsilon were flexible with tight schedules. It took us one month to get from sketch to a working application.

Marc RobitzcatGlobal Director of Marketing Technology
Marc Robitzcat, Global Director of Marketing Technology

The team at Appsilon has been a great complement to our own internal data science team. Their work has proven to be valuable in Diversey’s development of data-driven solutions to satisfy the demands of our global customers.

Sergey Fogelson, VP
Analytics and Measurement Sciences

We are big fans of your shiny.semantic package for Shiny and have used it to make our dashboards not ugly here at Viacom. Thanks for that!

Petter Arnestad, CEO
at Storekeeper AS

The Appsilon Data Science team took our Shiny app to a new level. They promptly delivered a new user-friendly interface, gave us a fresh new design and incorporated our logo and branded elements elegantly. Everything was delivered on time and within budget. We highly recommend Appsilon as a professional business partner in Shiny and R-related projects.

Triple A - Risk Finance
Frank Gregorkiewicz, Managing Partner, Triple A

The platform delivered by Appsilon team allowed us to migrate a specialized spreadsheet model into a collaborative web application and to give our clients unique tool enabling fulfillment of quantitative requirements of Solvency II Directive in effective and user-friendly manner.

NAMU banking
Thomas Ko, CEO
NAMU Systems

I would like to strongly underpin their major assets as professionals and co-workers: being extremely flexible and adaptive to the cutting-edge technologies combined with maintaining top quality of the code. Needless to say, we are deeply satisfied with the results of our cooperation.

Technologies we use

Amazon web services

Meet us

Load previous
November 7, 2016
@ EARL Boston

Effective applications of the r language.

June 5, 2017
@ EARL San Francisco

How we built a Shiny App for 700 users

May 15, 2018
@ eRum in Budapest

Taking inspirations from proven front-end frameworks to add to Shiny with 4 new packages

June 5, 2018
@ Predictive Analytics World in Las Vegas

Dynamic Pricing in real time for B2B Sales with R and Shiny

September 24, 2020
Why R?

This year, Why R? will be a virtual event.

November 9, 2020
PyData Warsaw

The PyData community gathers to discuss how best to apply Python tools (and R) to meet evolving data challenges.

January 24, 2021

Appsilon is a Full Service Certified Partner of RStudio and will be actively participating in next year’s virtual conference.

Load upcoming
PP-YOLO Object Detection Algorithm: Why It’s Faster than YOLOv4

PP-YOLO Object Detection Algorithm: Why It’s Faster than YOLOv4

PP-YOLO is a machine learning object detection framework based on the YOLO algorithm. In this article, we’ll explain what PP-YOLO is, why it is an improvement over YOLOv4, and show you how to use PP-YOLO for object detection.
Video: How To Set Up A Distributed Data Science Team

Video: How To Set Up A Distributed Data Science Team

In this video, Appsilon Senior Data Scientist Olga Mierzwa-Sulima explains best practices for data science teams – whether they are working in the office together or fully remote.
shiny.worker: Speed Up R Shiny Apps by Offloading Heavy Calculations

shiny.worker: Speed Up R Shiny Apps by Offloading Heavy Calculations

Long running calculations can freeze the UI of Shiny dashboards until the calculations are complete. Appsilon has created an R package to offload long running calculations so that the UI of Shiny apps can remain responsive.
Make R Shiny Dashboards Faster with updateInput, CSS, and JavaScript

Make R Shiny Dashboards Faster with updateInput, CSS, and JavaScript

This article will show you how to speed up R Shiny by pushing actions to the browser. You will learn how to omit the server bottleneck with updateInput, leverage CSS classes, and take advantage of JavaScript actions for smoother Shiny performance.
Are Computer Vision Models Vulnerable to Weight Poisoning Attacks?

Are Computer Vision Models Vulnerable to Weight Poisoning Attacks?

Recent research on weight poisoning attacks on pre-trained models in the area of NLP may transfer to the context of Computer Vision, where the usage of pre-trained weights is widespread.
Remote Data Science Team Best Practices: Scrum, GitHub, Docker, and More

Remote Data Science Team Best Practices: Scrum, GitHub, Docker, and More

Learn best practices for setting up a data science team and kicking off a data analytics project – remote or otherwise. I'll cover Scrum methodology, Asana, GitHub, Docker, renv, linter, and a variety of other tools and workflows.
Filip Stachura
Filip Stachura

Reach out. We’ll respond within 24 hours.