Engineered the high-performance tile-map server technology for the company that provided all the custom map overlays displayed by the universal map applications (above). Features included:
In 2010, at FyteComm, I used the PostgresSQL DBMS with the PostGIS extension for managing points, lines and regions where PostGIS could calculate region centroids and min/max extents, find line, point and region intersections, find the union of two regions, export region data as geoJSON, as-well-as speed up geo-queries by using optimized tree-search indexes. It was used in conjunction with customer data such as property tax data (i.e. owner, address, property lines, assessed value, zoning, etc). I just had to learn the PostGIS extension API combined with fairly standard SQL queries to use this database. I just did the normal single-server deployment pretty much out of the box and didn't worry too much about performance optimizations as our customer base wasn't that large.
Experiences using this skill are shown below:
Many of FlyteComm's customers had proprietary map overlays that they didn't want others to see. However, FlyteComm's array of virtualized map tile servers were shared among all customers. I updated my FastCGI tile server C++ code to authenticate every tile request using MD5 HMAC authentication using secret keys known only to the tile servers and ASP .NET app servers.