10 years of turning good ideas into great software
RX Security prints medical prescription pads for the North American market that meet the rigorous Department of Justice specifications. We built the software that powers their business. Almost a decade ago, we developed a desktop application to manage customer accounts and generate medical prescription pads (as well as printing plates destined for the print commercial printer) to be sent to doctors around North America. Recently they asked us to move their business to a cloud application in order to provide a better customer experience. This new web application will streamline their customer interaction, automate their product order process, provide better business intelligence tools and eventually allow for data analytics tools to help them grow their business in new markets.
Ontario Student Evaluator
Since 2016 we have been the technical team running product development for a studentevaluator tool used by teachers throughout Ontario. This application allows teachers to quickly generate and edit the reports that they have to submit on each student with eachreport card. We continue to extend this widely used product adding new features each year. By out sourcing the development to Colibri Software, the product owners can concentrate on building their business and know they have a reliable technology partner to build extensions, manage software patches, maintain their servers and keep their platform running.
Drug testing is big business especially in industries like trucking and manufacturing. It iscritical for companies that operate trucks that cross the US/Canadian boarder to keep accurate records of their drivers drug test results. In early 2018, we took over the development of the Veriport web application that provides a critical solution in that market. Over the last two years, we have managed development, production pushes as well as server management.
Azorus Inc. is a leading provider of student recruitment tools (The Azorus CRM) to universities in Europe and North America. Part of their offering is an analytics suite we built to meet their specifications. It extracts data from their CRM to be rendered in dash boards and processed through predictive analytics algorithms.
Bell Canada needed tools to help them predict customer churn, and developers from Colibri Software were part of the team that developed these tools. The suite of tools included ETL pipeline to load large portions of data in chunks into an analysis suite and use it to study the customer journey to predict when a customer might be considering leaving Bell. The team used state-of-the-art machine learning technologies to extract valuable insights from time-based input sequences. The infrastructure developed was capable of real-time analysis, allowing Bell to query our models on the fly amidst rapidly changing market conditions.The pipeline was built using Python, using Pandas and Numpy for data processing, and using Keras and Tensorflow for prediction models. Plotly and Matplotlib was used for data visualizations. The software integrated with Bell’s existing data infrastructure, and utilized GPU resources for model training.
US Medical Group
We worked with a large US medical group developing machine learning technology to automatically segment 3D MRI images into individual organs. We helped to architect and implement a solution that converted experimental scientific work into practical, scalable, production-ready software.
Fundy Language Analytics is looking to reshape second language acquisition by driving language instruction from the linguistic background of a particular learner. Fundy's tool set allows an instructor to set up maternal language based lessons that help the learner practice pronunciation problems that they are having. The system detects pronunciation errors and flags them for the learner. Most recently, we developed CTC (Connectionist Temporal Classification) models that help more accurately find phoneme boundaries in order to sharpen the results.
Our project with AT&T Labs was to provide the ability to monitor alarm conditions in over 200 labs across North America from a single web interface. Included was a dashboard so that alarms status, specific resolution team details as well as alarm resolution could be tracked and reported to senior management. To achieve that goal, we developed a distributed system composed of many “agent” computers (low power Linux computers) monitoring sensor data (one in each lab facility)that communicated with the building’s BACNet stack to acquire data from the Siemens BAS (Building Automation System) and post it securely over SSL to the cloud API.These agents feature user configurable interface, a management dashboard to monitor and manage the status of these agents, a “dispatcher web interface” used to dispatch a North America wide distributed team of technicians when problems are detected in a lab, as well as a cloud API to collect the data acquired by the agent in a secure manner. Finally, we developed a NOC application to manage all the agents across multiple sites.
Our initial project was to set up a responsive app to connect to an SNMP enabled device, acquire data from the device by “walking the data tree” and displaying that data in a rudimentary form. This presented a number of security challenges since the SNMP protocol made security(user authentication) an issue. We developed a cloud application that could establish a secure connection to the devices and provide the appropriate user authentication that the legacy devices themselves could not. The user was then able to access that data from a mobile device at all times.The success of that project lead to a full scale multi year project to build interfaces for the next generation transmitters.
Community Based Monitoring Program (Nunavut Wildlife Management Board)
Noreca Consulting works extensively in Nunavut to understand indigenous needs and provide technical solutions. In 2011, Noreca approached us to develop a mobile app to record data collected in the field by hunters and trappers, using the Juniper Systems, Mesa device (a rugged Windows phone). The application was written to support multiplelanguages, including English, Inuktitut and Inuinnaqtun.The application tracks a hunter's trip by GPS positioning and then records animal observations or specifics on any animals harvested. A data file for each trip is uploaded to a web server which records the state of this critical resource for remote communities of Nunavut. We participated in training in Nunavut in January 2012.
Growing out of initial work we did on “context aware” content management, we developed a geo-located CMS in the form of an API that manages content tied to a specific point on the earth. The backend tracks how users interact with the content and manages comments that they have at any point. The American Geographic Society recognized our early efforts to provide a community based tool to map small geographies and provide rich location based content to visitors wishing to explore on their own.We have launched custom apps for towns, outdoor clubs, historical societies or other organizations that have outdoor assets that they would like to share. We produced a short YouTube video (https://www.youtube.com/watch?v=nrTTsy0DPzk)about the Strollopia experience.