Alexander Makeev

I implemented the full-stack architecture for this wedding-locator service. Technologies used in development were:

Server-side: Nginx, PostgreSQL, SQLite, Python, Django, Python Imaging Library (PIL/PILLOW)

Client-side: JavaScript, JQuery, AJAX, Handlebars, HTML, CSS

I implemented the full-stack architecture for this service. Technologies used for development were:

Server-side: Nginx, PostgreSQL, SQLite, Python, Django, Python Imaging Library (PIL/PILLOW), Node.js, Grunt.js, Handlebars Precompiler, PHP

Client-side: JavaScript, JQuery, AJAX, Handlebars, HTML, CSS

Designed and implemented full stack architecture for these fifteen landing pages (pixel-perfect HTML+CSS was implemented by another person).
Technologies used for development were:
- Server-side: Nginx, Node.js, Grunt.js, Handlebars Precompiler, PHP
- Client-side: JavaScript, JQuery, AJAX, Handlebars, HTML, CSS

This project was based on parsed US government insurance plans databases. It included a flexible filtering and adaptive cost estimation algorithm (factored in age, smoking factor, annual income, family members, and automatically calculated subsidies).

Course editor allows the creator to construct educational steps with embedded videos, illustration images, formatted texts, and formulas.

Courses portal includes flexible course access management (grouped access lists with tagged resources linkage) and running courses, with progress tracking included as well.

Detection engine is based on minutia points detection. Web-UI allows users to add new fingerprints and search the entire database by fingerprint images.

My work on this project included: automated email notifications, status tracking, and specialized tools like template insurance offering.

The tool helps web developers to validate JSON code, layout it in tree-based form, and navigate through collapsing tree branches.

Also, the tool supports simple installation to command line for Debian-based operating systems and MacOS; users can use it on any local JSON script output, sending it automatically to a dynamically opened browser tab.

Online color filter tool for checking that a website is suitable for color blind persons. URLs are filtered for multiple kinds of forbidden content, then dynamically rendered as an image with a color-scheme generated according to color blind modeling algorithms. Multiple color blind algorithms implemented:
- Protanopia (red/green color blindness; anomalous red cones)
- Deutanopia (red/green color blindness; anomalous green cones)
- Tritanopia (blue/yellow color blindness; anomalous blue cones)
- Greyscale/achromatopsia (quick check for all forms of colorblindness)

This tool optimizes the sprite loading technique that websites use. A user can drag-and-drop or select sprites from local storage and get them rendered as a single sprite container image with CSS snippets automatically generated.

This system allows to users specify custom questions to help distinguish client state. More than 15 websites are implemented on the same core system with specific questions and logic.
The system supports flexible configuration of question lists, answers variants, answer logic, and results estimation logic, with dynamic report generation of why the system gave a certain result.

All data is dynamically analyzed and displayed as multiple layers as a specifically configured Google Maps widget. Zip-code regions are highlighted with detailed stats and hovering descriptions. Multiple tools with different analysis algorithms have been implemented.

Management tool designed for databases of at least 6 million data records (engine tested to perform well for up to 20 million records). Dedicated storage server configured with optimal storage costs and management efficiency. Automatic conflicts resolver implemented with the flexibility to add further rules and resolve scenarios adding.