Mar-Apr 2020: Initial Postgresql migration, stalledDec 2021-Mar 2022: (Finalized) Database migration, from MSSQL to Postgresql. Migrated complex stored procedures hundreds of lines long. Learned some Python, Scala, Java, and Rust while migrating a data ingestion script originally written in Perl. (It was working in Python but slow, so I rewrote in Scala for practice, but I didn't like the large .jar so I rewrote as a Java class and some ColdFusion code, which is the currently deployed solution. However, I practiced rewriting it in Rust partly as a speed benchmark, and partly to understand the language better.)Jul 2022 - AWS Migration [On hold, but used as a backup]- Migrated 3 years of Postgresql data to a free instance of RDS (1GB) and a web server (also 1GB RAM). Learned AWS root user management, IAM and security group configuration, EC2 VM creation and administration, use of Amazon Linux 2, RDS, S3, CloudWatch, EBS administration, backup configuration (for the web instance and for RDS), configuration of firewall rules for FTP (Proftpd) (active and passive), use of AWS command-line tools.Mar-Apr 2023: POS Integration project - ColdFusion(CFML)+HTML collects OAuth tokens from stores that use LightSpeed- Python: 1) LightSpeed to JSON, 2) JSON to a Streetpulse import format.- Rewrote in Rust: Combined extract and convert into one program, with token refresh and retry functionality. (Much time was spent learning Rust error-handling.)

- Sep-Nov (MMSALES - MMC Manual Reporting): MS was changing accounting procedures to an accrual model. There was a need for some ad-hoc queries to fill in for some missing reports. Created a few SQL queries based on sparse specifications (field lists). Spent a bit much time tweaking my mock-up data, and spent far too long (about 3 weeks) studying about 15k+ lines of SQL, while waiting for good data.

- Added visualizations to Power BI dashboards using datasets coming from Azure Data Explorer and Power Query. Used M/web services for some data.

- BI/Reporting: Optimized a Power BI dashboard by converting Tabular to import model. Semi-automated this process with an F# script.- Automation: Created an F\# app to collect query and connection information. Created F\# and Powershell scripts to unprotect spreadsheets

- BI/cube work: Rewrote/fixed/productionized ETL for a cube (using SQL MERGE instead of SSIS), then cleaned up the cube, removing some redundantdimensions.- Web scraper/Power BI dashboard for click-through rate metrics: Scraped code off a (Windows) hard drive using Linux. Optimized SQL. Converted C\# code to F\# and added archiving of scraped pages to Sqlite for quicker testing of the html-parsing logic.

Developed and maintained two of Caradigm's multidimensional databases (cubes) in SSAS & MDX. Ensured that MDX queries met the 2 second SLA, by moving all WITH clauses at the top of the query into the cube calculation script, and by optimizing the design of the Claims cube (no LastNonEmpty / versioning of facts.) (Claims cube had much better performance.)Wrote some F# and Powershell scripts for benchmarking and validation.

BI POC, successfully delivered to client. (Some details: Custom rollup operator and MDX working. Currency conversions working. Materialized the currency conversions for "Movement"; using LastChild aggregation for "Accumulation". Used relationships between dimensions and currency exchange measure groups to keep the MDX subcube assignments (cell calculations) simple. Delivered several F# notebooks to the client as examples of how to semi-automate use-case tests, performance tests, and partitioning.)

SSRS, MS SQL, C#, F#1) SSRS: Created reports for a trade-recording application.2) SSRS: Converted Crystal reports for the "CR" application.3) C#: Created a web-service downloader with XML config4) F#: SSRSDeploy.fs, SSRS downloader, SearchReplace.fs5) SQL/SSRS: Resolved various reporting tickets for Global Indexes group, involving some complicated SQL at times.7 week break in March/April, started again in May8) More SSRS reports, this time with a Web API service behind it. Most of the time has been spent on the Web API, and SQL triggers.

On assignment at AT&T, using Oracle (TOAD) for adhoc queries, Perl for scheduled reports• Responded to customers' ticket requests• Created a generic query-to-spreadsheet-to-email Perl script• Automated recurring queries• Set up SVN server and clients• Instructed coworkers on check-in procedures for deployment of reports

• SQL: histogram procedures for all columns of all tables (for comparing databases)• SQL: SQL to SQL population based on a variety of lookup tables, with retry logic• SQL: DB2 to SQL population, adjusting for row limit and time limits• SQL: Optimized CRC32 from 1600 microseconds. to 8, then to 0.25 in F#• C#: TSV import, including a regex technique to fix for extra tabs• F#: SQL count comparison; compares tables showing counts of differing values for each column• F#: Population testing, builds the databases and compares old and new techniques

• Developed some of an FTP to AWS S3 utility.• C++: fast C count validation program based on simple hash• SQL/SSAS/SSIS/VS2008: Migrated a client's database & cubes from the MSFT data center to AWS. Partially migrated a second client (optimized 5 hrs query to 7 mins by adding an appropriate index).

• SQL/SSRS: Developed site metrics reports. Used stored procedures, recursive queries (CTEs), and some xml functions.. • SSIS: Imported web logs for reporting, while parsing out GUIDs into a separate column.

Data warehousing and reporting.SQL, CFML, Perl, ASP, C, VBS, .NET development• Maintained and developed a data-warehouse-like application written in ColdFusion:• SQL: Tuned queries and indexes for fast reporting. • SQL: Built an hourly-updated, ROLAP-like table of aggregates with ranks, for the “market title” report. • Perl: Rewrote the import system. • SQL/CFML: Reduced redundant code (generalized sp_getsales & report pages). • SQL: Implemented hourly log backup, nightly database backup and nightly updates to various aggregation tables. • SQL: Implemented report caching to increase throughput. • Various (Cygwin): Setup Cygwin, SSH, Unison files synchronizer, Perl (cpan modules) for maintenance of the site, scheduled tasks, batch files, bash scripts. • SQL: Simplified the permission system. • Perl: Automated weekly product update and sales backfill (http download). • Perl: feeds (daily exports), “pinger” to alert for availability problems (HTTP/FTP).• Excel w/SQL: Maintain a small product list. Make data changes as necessary for occasional maintenance. • Hardware/Linux+KVM/Xen/Hyper-V: Built and migrated servers six times. • Perl: Set up a replicated backup site using custom-built replication to minimize bandwidth. • SQL: Migrated to SQL server 2005. • SSAS/MDX: Built OLAP cubes for eventual migration to OLAP. • C++ ATL (Win32 native console): SPUpload: resident upload client with FTP and SOAP methods• ASP.NET+VB: upload web services, clients in VBScript, Perl, C#• C# (Win32 forms): SPScan: store UPCs in SqlLite, upload only newer data to FTP or web service2008:• VS/C#/MDX/SQLCLR: ASP.NET MVC markettitle report demo w/custom MDX denseRank C# CLR user-defined function• Maintenance, ad-hoc queries/reports, backups, migration, virtualization, recovery

• Maintained Bibliofind (2 years), including maintenance of a Access credit-card billing application, answering customer-support issues, and making script changes as needed. • C++: Developed part of an import processing application for importing Bibliofind’s customer records to zShops. • C++: Developed an inventory control application for a partner. • C++/Oracle: Fixed memory leaks and identified query inefficiencies.• Bash/Perl: Wrote scripts to archive/delete old files, monitor & bounce failing java processes.

• Migrated Bibliofind to new servers• Answered customer support issues including some ad-hoc queries and script debugging• Made some modifications for security concerns.

• Wrote C applications, TEXIS webscripts (CGI), and HTML for a website used by millions (over 100,000 visitors per day at peak). • C: Implemented importing and email in C++, using scheduled nightly batch files to switch to a newly indexed database once per night.• C: Implemented “want matching” first in SQL 6.5 then in a C++ app written using Thunderstone’s TEXIS search engine.• TEXIS webscript: Implemented search, shopping basket and emailing pages• Performed all my own testing on Bibliofind scripts and applications.

• Visual Basic: created a “report viewer”, helped with a bug in a communications app.• Excel: Created an accounting workbook with posting macros• Excel: Helped pull data from SQL into a spreadsheet for a utility cooperative.• VB/C++/Texis webscript: Start of work on Bibliofind, including initial import program, search, and initial “want matching” engine in MSSQL.

• Excel: Data entry, price list management, sales reporting, charting, budgeting• Rumba/Access: Downloaded from AS/400, fixed bad separators with SNR, and imported to Access to do some departmental reporting without contacting the IT department.