Gradient Healthcare is a Shenzhen-based digital medicine startup with AI applications. Our advanced AI algorithms abstract information from daily health data collected from the wearable devices and provide seamless personal care through quantitative analysis.

Alumni of Canadian Universities (ACU) is a young and dynamic organization founded in 2017 by the official alumni reps of University of Toronto, McGill University, University of British Columbia and University of Alberta.In 2019 ACU is able to reach out to 1,800+ alumni by connecting more Canadian university alumni groups in Southern China.ACU is aiming to provide an open-minded platform for Canadian university alumni to share, to connect and to create.

Semacare is a Melbourne/Shenzhen startup that focuses on innovation in the medical field. The main business comprises medical grade hardware products such as the single lead ECG monitoring device and e-skin ECG monitoring patch (single & multiple leads), and software services like arrhythmia detection and preventive algorithmic services based on clinical data. The management team comes from Mindray, Baidu, and Sensetime.My responsibility includes:On the corporate level-… Show more Semacare is a Melbourne/Shenzhen startup that focuses on innovation in the medical field. The main business comprises medical grade hardware products such as the single lead ECG monitoring device and e-skin ECG monitoring patch (single & multiple leads), and software services like arrhythmia detection and preventive algorithmic services based on clinical data. The management team comes from Mindray, Baidu, and Sensetime.My responsibility includes:On the corporate level- Participate in the corporate strategy and product decision making.- Create, hire and manage the research and development team (Product, Software, Data and Hardware).- Present the company in financing process- Be responsible for the technical part of due diligence.On the project level- Manage the negotiation and cooperation with Chinese medical institutions.- Administer the e-Skin product realization and mass production from the cooperation with the Monash University and the NanoBionics Lab.- Lead/Search the cooperation with universities on the application of medical big data.On the team level- Formulate the workflow, the tools and the technical stack of the development team.- Analyze and document the product requirement for the software, data and hardware products.- Do all the dirty work of the backend/server side.- Set up the hardware and build the stack for the deep learning of the medical big data (ECG).- Preprocess the data and train the models.- Supervise the development and the manufacturing of the hardware products. Show less

The Misfit app and a group of portfolio apps are used to work with Misfit’s Flagship products and partner products (e.g Speedo and Swarovski). More than 2 million paid users dwell on the apps, which in turn need to support syncing with a dozen or so hardware products, present data for different purposes and create unique experiences. My work is composed of - scripting annual as well as each sprint’s plan following Misfit/Fossil strategies and demands from diverse aspects of the… Show more The Misfit app and a group of portfolio apps are used to work with Misfit’s Flagship products and partner products (e.g Speedo and Swarovski). More than 2 million paid users dwell on the apps, which in turn need to support syncing with a dozen or so hardware products, present data for different purposes and create unique experiences. My work is composed of - scripting annual as well as each sprint’s plan following Misfit/Fossil strategies and demands from diverse aspects of the company and partners, - keeping smooth and efficacious communication across teams and regions, conducting regular collaboration with the e-commerce/marketing/business development/customer service teams, - having product managers draft product requirement documents and appropriately executing them, - holding daily scrum, kickoff and sprint review meetings, - working closely with engineering teams so that they could deliver apps on time, - from time to time joining forces to create new hardware products.From the time I taking over the ship, we have - solved more than 1.2K legacy bugs in the first two months, - lifted our rating on iTunes from the quagmire of negative feedbacks to a stable 4 star (out of 5) - Android is more complex but we see increasing climbing in rating,- straightened out our relationships with the customer service team (Vietnam), the firmware team (US/VN), the QA team (China) and the design team (US) by setting up and, strictly following and regularly reviewing cross-team agreements and workflows, and holding weekly online meetings, - become a reliable companion to other teams and partners with expected shipping deadlines and fast feature development. To promote the value of software in the ecosystem of a hardware company, I have also - adopted a content-based new strategy to attract new users and improving our retention rate,- proposed the firmware upgrade premium service to consolidate our user service and activating collaboration with partners (Speedo). Show less

Previously we didn’t have a serious data team to support product design, business data analytics and data-driven engineering requirement. My responsibility includes - collecting requirement from different teams,- building the data team from scratch- pitching data services to other teams, - driving other teams to open source their data,- designing the data pipeline, ETL and BI architecture, - creating the dimensional model, - developing scripts on EMR Spark… Show more Previously we didn’t have a serious data team to support product design, business data analytics and data-driven engineering requirement. My responsibility includes - collecting requirement from different teams,- building the data team from scratch- pitching data services to other teams, - driving other teams to open source their data,- designing the data pipeline, ETL and BI architecture, - creating the dimensional model, - developing scripts on EMR Spark for ETL (Extract, Transform and Load), - automating data pipelining with Airbnb Airflow- conducting services for commercial, engineering and customer service teams.I structured and developed the data pipeline and analytics platform (BI) that - extract application data from MongoDB, user behavioral data from Flurry and Amazon Mobile Analytics as well as bluetooth sync log and device hardware log, - transform them into unified data schema, - load the processed data into Redshift,- support internal users to query the big data with Periscope,- work with data scientists and provide actionable insights from user's activity and sleep data. Show less

Home is the app (Android & iOS) for Misfit’s first Smart Home product Misfit Bolt. Bolt is a Bluetooth smart bulb that supports direct connection with the phone (without a Philips Hue-like hub), various dynamic scenes, turning on/off the bulb with proximity information, sunrise feature, etc. My responsibility includes - project planning, - product feature requirement drafting and management, - sprint management, team management, - maintaining efficient communication… Show more Home is the app (Android & iOS) for Misfit’s first Smart Home product Misfit Bolt. Bolt is a Bluetooth smart bulb that supports direct connection with the phone (without a Philips Hue-like hub), various dynamic scenes, turning on/off the bulb with proximity information, sunrise feature, etc. My responsibility includes - project planning, - product feature requirement drafting and management, - sprint management, team management, - maintaining efficient communication among different offices in the US (Product), Vietnam (Firmware and SDK) and China (Development and Manufacturing). The results of my work include - a successful team reorganizing, - long-lasting sync issue getting well alleviated,- innumerable legacy bugs fixed, - promoting a series of automation in testing (BDD/Stability/Performance) and in data reporting, - a heavy reliance on Jira for bug tracking and Phabricator for code review- the team becoming a reliable partner among offices. Show less

Project 1 (2014.6 - Present): Simeji for iOSDescription: Simeji is the most popular Japanese Input Method for iOS 7/8. I took part in the system design and implemented the Extension’s top bar, options, skin change, keyboard change as well as its containing app’s interface and custom dictionary functions.Project 2 (2013.1 - 2014.5): Baidu Browser (Android) for the International MarketI was in charge of tech team management and backend development. My technical… Show more Project 1 (2014.6 - Present): Simeji for iOSDescription: Simeji is the most popular Japanese Input Method for iOS 7/8. I took part in the system design and implemented the Extension’s top bar, options, skin change, keyboard change as well as its containing app’s interface and custom dictionary functions.Project 2 (2013.1 - 2014.5): Baidu Browser (Android) for the International MarketI was in charge of tech team management and backend development. My technical duty includes design and implementation for a million throughput/daily system, a multilingual news page transcoder, a text summariser based on sentence similarity and a set of service monitoring tools. In addition, I introduced the use of Jenkins to automate daily build, Sonar to supervise code quality and Fabric to regulate code deployment. Show less

Project: SearchetDescription: The product leveraged OCR-like technology to enable two-step search for texts and images presented in arbitrary apps. A team of five people have worked on it. We received seven Chinese patents in total. Video of the running product would be available upon request.

Project: Socrates ReaderDescription: Apart from the initial bewilderment of several months I found after joining an IT corporation, I proposed Socrates Reader which was designed an intelligent app that grabs from users’ twitter time line and the late Google Reader, and recommends articles using algorithms like LDA. My jobs: I implemented the backend, which is composed of scraping, processing (Chinese segmentation and data cleaning/storing), modeling (LDA & Naïve Bayes), ranking,… Show more Project: Socrates ReaderDescription: Apart from the initial bewilderment of several months I found after joining an IT corporation, I proposed Socrates Reader which was designed an intelligent app that grabs from users’ twitter time line and the late Google Reader, and recommends articles using algorithms like LDA. My jobs: I implemented the backend, which is composed of scraping, processing (Chinese segmentation and data cleaning/storing), modeling (LDA & Naïve Bayes), ranking, layout (mainly Readability code), and a Flask-based REST API. Show less

Project: Integrating across Space, Time & Gender in the Humanities for Chinese Literary, Historical & Geographical DatabasesDescription: ICHD was funded by SSHRC, joint project of McGill, Harvard and Peking Universities. I was supervised by Prof. Renée Sieber. My jobs were to create Upper Level Ontology and domain ontologies using Protégé and OWL, designed the structure to support SPARQL query in three semantically different data sources (persisted in MySQL and via Web… Show more Project: Integrating across Space, Time & Gender in the Humanities for Chinese Literary, Historical & Geographical DatabasesDescription: ICHD was funded by SSHRC, joint project of McGill, Harvard and Peking Universities. I was supervised by Prof. Renée Sieber. My jobs were to create Upper Level Ontology and domain ontologies using Protégé and OWL, designed the structure to support SPARQL query in three semantically different data sources (persisted in MySQL and via Web Services), wrote a parser to translate SPARQL into REST requests and integrated different parsers into D2RQ. Show less

Project: SRed: A Secure Reputation-based Dynamic Window Scheme for Disruption Tolerant Networks (DTN). Description: This was a conference paper project led by Dr. Zhong Xu. My technical liabilities include, implementation of the new algorithm on the simulator ONE (team of 1 developer), creating a new tolerant network map of Montreal, Canada via Google Maps (KML), creating a demo Android app on emulator (Android was then not released, team of 1 developer)

I was supervised by Dr. Yoshioka Nobukazu. The idea is first to study SecTro II that provides a complete involvement of Security Patterns, design patterns for ideally creating secure software from the Design Phase. The Common Criteria is then added to SecTro II as a plug-in in order to serve the industry demands.

Project: WSN Node Transmitting Range OptimizerDescription: I was supervised by Dr. Kumar Padmanabh. The idea is to determine each node's optimal transmitting range in the Wireless Sensor Network (WSN) by an algorithm optimised from the Genetic Algorithm.