# Interview Presentation by Guido Stein 2022-06-14 Note: Thank you for joining me today. This presentation shares information about me and my career in order to give you a better picture of who I am and how I think.

# Overview - Introduction - Example Project - Geospatial Input - Tabular Input - Workflow Design - Conclusions

# Introduction ## Guido Stein ### [gee-doh] like Burrito Note: I prefer my name pronounced [gee-doh] like burrito, mosquito, and Dorrito. That said, it's not that big a deal to me.

# Interrupt me Note: I want you to get everything you want out of this presentation, so please feel free to interrupt me if you have a question, comment, or would like more details.

# Highlight Reel - History - Proficiencies - Tools Used - Accomplishments - One More Thing Note: I have many qualities that make me a great candidate for Data Engineer. Here is a highlight reel of me and my career.

# History - B.A. in Geography from Clark University - 20+ years working with geospatial data, workflows, and development Note: I have a solid foundation in using geospatial data to solve problems. I started my geospatial learning at university and worked in the industry ever since.

# Proficiencies - Python - Javascript - SQL - Workflow Design - Problem Solving Note: I have mastery and understanding of the fundamental skills needed to build technical solutions.

# Tools Used - Node - Jupyter Notebooks - Postgres - AWS - QGIS Note: I have used many of the current tools, applications, and frameworks for developing solutions.

# Accomplishments - Reputation for geospatial data domain knowledge - Reputation as a technical problem solver Note: I am known for problem-solving with geospatial and other technical skills. Colleagues often reach out to me to help them solve issues.

# One More Thing - Co-Chair FOSS4G Boston 2017 - President OSGeo US Local Chapter - Co-Chair FedGeoDay Note: I am a well rounded and valuable team member who can execute, lead, and develop esprit de corps. I keep current with technological trends and practices by participating in the open source geosptial community.

# Example Project ## Mapping Access Note: To demonstrate how I work, I will share with you this example mapping project that I worked on for five years.

# Statement Of Work Create a reusable solution to collect, convert, and aggregate spatial data working with a team across two offices that report the output data multiple times a year for five years. Note: In short, we were collecting data about access to services across the U.S. with multiple service providers submitting data

# Variables - stake holders - proprietary information - data providers - team members - output formats - reporting dates Note: Each variable in this project is a multiplier on the number issues the solution being built needed to satisfy in order to be succesful

# Reporting Standards - Tiger block polygons for urban areas - Tiger road polylines for rural areas Note: Aggregated data needed to be reported in the Census Tiger format which is a common standard for U.S. demographics. It is both a blessing and a curse to use this format because there is variations from decade to decade because block boundaries and road networks change.

# Conflation The merging of two or more sets of information, texts, ideas, etc. into one. -Google Dictionary Note: Combining data is tricky because you risk the chance of mis-representing the source information in your output.

# Geospatial Input Data - Point data representing homes with access - Polygons representing areas with access - Lines representing roads with access Note: Some of the service providers submitted spatial data in the following formats.

# Issue #1 ## Road to Road Conflation  Note: Data providers used multiple reference system geometries for the road linework. Alignment of geometry or 1:1 relationship caused issues.

# Issue #1 ## Solutions - Buffer provided segment and then join to Tiger segments that are within buffer - Snap provided segment end points or vertices to Tiger segments then join overlapping geometries - Convert provided segments to center points and then join to nearest Tiger segment

# Issue #1 ## Reasoning and Research - Buffering segments does not work if new segments along a road are smaller than the potentially associated Tiger geometry - There is no way to set a proper tolerance for snapping to Tiger geometry - Center points are not exact, but they can represent 1:* relationships and can assign data from new roads to nearest segment in Tiger geometry

# Tabular Input Data - x,y representing homes - census tiger ids

# Issue #2 ## Processing x,y points  Note: x,y point data commonly uses decimal degree latitude and longitude coordinates and Tiger Data is in NAD 83 feet.

# Issue #2 ## Solutions - Convert all incoming data into NAD83 before processing - Convert all reference data into WGS83 before processing

# Issue #2 ## Reasoning and Research - Each provider data set sent will need to be converted to NAD83, making this a time sink - Only the output data will need the original source geometry from Tiger - None of the provider data we looked at came in as NAD83

# Workflow Design 

# Issue #3 ## What tools should I use?

# Issue #3 ## Workflow Design Goals - modular tools - self documenting process - accessible for GIS Analysts

# Issue #3.1.0 ## Use FME Desktop - Develop reusable processing models quickly - Conversion models act as documentation - Limited licenses available

# Issue #3.2.0 ## Use ArcGIS Data Interoperability extension - Floating licenses available - Toolset accessible through familiar ArcGIS - Editing environment buggy and loses model updates

# Issue #3.3.0 ## Use ArcGIS Model Builder - Can build reusable models - Models are self documenting and can be stored with processed data - Models get too complicated and fail when using variables

# Issue #3.3.1 ## Use ArcGIS Arcpy Scripts - Can build reusable scripts - Not self documenting when run - Not familiar to GIS Analysts

# Issue #3.3.2 ## Use ArcGIS Model Builder and Arcpy Scripts Tools - Can build reusable scripts - Models add self documentation and are stored with processing data - GIS Analysts can use Model Builder to access Python scripts

# Issue #3.3.2 ## Results - Time spent processing data improved with each round by at least 25% - Documentation of process allowed for auditing providers for QC - Modular design allowed for new logic and processes to be added to the workflow when needed

# Conclusion - I have a deep understanding of geospatial concepts - I have strong skills and knowledge of multiple types of data and technology - I enjoy developing, building, and troubleshooting data workflows - I don't mind if you call me Gwee-Do Note: I hope that you take away the following from this presentation

# But why REsurety, Inc.? - Human generated climate change is bad - You work with tools I like to use (Python,AWS,Postgres) - You are building and iterating solutions - Interviewing has been joyful and makes me feel like there is a good culture fit

# ?? Questions ?? ## Interview Presentation by Guido Stein 2022-06-13