 Responsible for team developing algorithms to help concussion patients recover.  Apply machine learning and statistical analysis on EEG and other physiological signals.  Evaluation and recommendations of EEG hardware.  Responsible for IP portfolio.  Advise on user experience. Advise on research protocols and data collection.

Applied machine learning, statistics, exploratory data analysis to build prediction models for government, manufacturing and trucking logistics industry. Notable projects and accomplishments include:- Developed Natural Language Processing proof-of-concept to help health care workers get Government of Canada approved vaccine and immunization information.- Developed prediction models of downtime for a paint line schedule for a Tier 1 producer of automotive parts.- Built prediction model of estimated time of arrival for trucking logistics company using Pyspark in Databricks.

Lead Strata’s Data Science team to determine the financial contribution of brand and brand ROI by analyzing any and all organizational, customer and industry data that impacts brand. Also lead the brand strategy and investment modeling to provide optimization scenarios based on clients’ business objectives, budgets and constraints so clients can understand the organizational impact of brand investments. Using Strata’s Brand Economics™ model, we measure the entire acquisition/conversion journey to understand which strategic and tactical levers will produce the optimal business outcomes.

End to end pipeline and algorithm development for a wearable watch measuring heart signals.
Developed algorithms in Python and Matlab that included: stress/anxiety prediction, blood pressure, heart rate variability based biofeedback, physical fitness and breathing rate.
Statistical analysis and Digital Signal Processing were applied to raw data stored in Google Firebase cloud to learn parameters and evaluate algorithm accuracy using Python and Matlab.
Coded stress prediction and HRV based biofeedback algorithms in C++ for integration into smartwatch app.

Conducted R&D of algorithms for analyzing and classifying brain states; used digital signal processing with machine learning techniques for Muse dry electrode portable headband EEG system. Mined EEG and demographic data in Cloud. Researched predictive models of drowsiness, concentration and facial muscle tension. Oversaw team of internal and external researchers. Managed Intellectual Property portfolio, as Project Lead in filing utility/design patent applications; facilitated cross-functional brainstorming sessions, wrote detailed technical descriptions (~100 pages), collaborated with patent attorneys to file applications/respond to actions raised by patent examiners. Assisted in composing grant proposals. Machine Learning and HCI R&D- Contributed to Muse product launch and supported ongoing research on new product applications by spearheading research initiatives and seamlessly managing multiple detailed processes.- Produced real-time audio user feedback by designing algorithm for Muse mobile app. Led development of Cloud application. Created algorithm that uses Cloud infrastructure to determine brain models.- Paved road for entry into medical market by demonstrating substantial improvement in anxiety scores across hundreds of users after eight weeks of product use. - Maximized signal quality for headband with no assistance on first try for 90% of Muse users. Teamed with User Experience designer to design real-time indicator to maximize signal quality and minimize frustration.- Enabled largest repository of EEG data in the world, big data stored in Cloud, with largest ever study of its kind across tens of thousands users; published in peer reviewed article. Wrote Matlab code; developed SQL interface. Patents - Secured utility patent and dozens of design patent grants across Japan, China, US, Korea and Canada. - Completed six patent applications as inventor. - Gained broad protection of technology by including hardware, algorithms and Cloud technology.

- Funded by the Alzheimer’s Society Canada to develop and validate the electronic Montreal Cognitive Assessment (eMoCA), a cognitive screening test designed to assist health professionals in detecting early changes in cognitive function.- Worked with HP Canada and large Canadian hospital to co-develop and evaluate health care applications using early stage technology being developed by HP Labs.- Co-authored white paper (Telus Health) with vision for industry to follow to 2020. - Examined barriers and drivers for the adoption of an electronic patient tracking board (McKesson Performance Visibility) by a large Canadian hospital; findings will be used in marketing.

Applied traditional biostatistics and advanced machine learning techniques to the analysis of stroke and dementia that searched across hundreds of variables from the Canadian Study of Health and Aging. - Awarded Alzheimer Society Canada Post-Doctoral award of $81,000 over two years.

Developed and evaluated a decision support system that uses data-mining and pattern recognition methods to discover knowledge among co-occurrences of feature values from data. Determined significant rules that improve classification and decision support. Major features of this artificial intelligence system include:- Transparency: system is capable of explaining the basis of its decisions.- Accuracy: system capably produced accurate decisions and is resistant to data “noise.”- Reporting: system provides a calibrated report of decision confidence or certainty.- Versatile: system simultaneously handles mixed mode data: discrete, continuous, categorical, and descriptive; this data mining tool’s methodology is highly applicable to business scenarios including marketing, risk and operational analysis, for strategic business decisions.

Patents:- Granted a patent video in the Fiber-to-the-User Project.- Granted three U.S. patents for LNP inventions.Accomplishments:- Contributed significantly to Nortel’s win of $1B in sales with outcome of complex LNP project. Travelled North America, met with over 20 telephone companies to conduct needs analysis; created economic model; played major role in setting LNP standards. Received President’s Award.- Developed a network architecture and design for a nation-wide network of 50 cities for internet and unified email messaging applications. Design contributed to win of $30M contract with Microsoft Network (MSN).- Managed development of a new architecture and model to determine capacity of metropolitan optical Ethernet network to carry forecast traffic loads that resulted in a $15M contract with Bell Canada.- Generated $90M in sales by writing industry standard specification Operator Services Number Portability.

Manager, Advanced Intelligent Network 1993-1994Developed Advanced Intelligent Network (AIN) switch requirements and AIN based telephony services.Manager, Switching Systems Technology 1989-1993Developed requirements, negotiated with vendors for switching platforms, peripherals, interface and remotes.Accomplishments:- Saved $2M in custom development fees, and further kept ongoing operational expenses low; led team that solved Call Redirect Service issues in-house.- Created contract based process that achieved a dramatic improvement in service development time still in use by Bell Canada today.

Technical sales representative to Nortel Canada & Bell Northern Research responsible for design wins for Mitsubishi Semiconductor components including microprocessors, laser diodes for fiber-optic transmission and D-RAM memory chips.

Digital PBX Product Engineering Design authority for digital Private Branch Exchange (PBX) technology: DS1 trunks, Basic Rate Interface (BRI) ISDN line cards, analog lines, and trunks.

Microelectronics EngineerManufacturing engineer and design authority responsible for thick film hybrid telecommunications circuits for digital, analog lines and trunks.

VLSI design of IBM PC compatible logic that integrated numerous MSI chips into a single package to reduce power, space and failure rates.