Consulting to corporate customers on multi-year contracts. Work included:• Provisioning, installation, monitoring and management of Frame Relay and C-Band Satellite networks across North America• Provisioning, installation, and configuration of ISDN services for trunking and fractional DS1 use for data, voice, and fax services• Document Management Services for legal, industrial, and logistics customers• Distributed Document Imaging, OCR, and Cataloging for legal, industrial, and logistics customers• Accounting rollup reporting for multiple accounting systems, charts of accounts, and accounting software packages for small-medium businesses up to multi-corporation IPO roll-ups

● Managed all AWS accounts and planned the migration from IAM anchored accounts to combined into an AWS Organization and AWS IAM Identity Center SSO managed constellation.● Supplied the initial Terraform configurations of all existing resources, and prepared BitBucket Pipelines for automated CI/CD deployment and drift detection.● Defined custom python-based AI Vector DB deployments and batch document ingest pipelines using S3 trigger-based updates.● Developed several PoC Large Language Models (LLM) using Langchain, HuggingFace and OpenAI APIs for inferencing integration with Amazon Chime, AWS Connect, Genesys Cloud CX and Google Dialogflow ES and CX. These deployments also included Amazon Lex and Polly chatbots.● Worked on several PoCs using AWS Bedrock LLMs for integration with the call center solutions in production.● Tools used: AWS Bedrock, Terraform, BitBucket Pipelines, various embedding and vector stores, Langchain, HuggingFace, OpenAI, Google AI Studio

● Air-gapped Python Data Quality Engine implementation, using approved Pydantic and SQLAlchemy libraries, and factored into a SOLID / hexagonal architecture to support future business requirements and delivered via GitLab CI/CD pipeline into Nexus deployment repos. The system integrated with existing homegrown ETL process as a pip installable Python library and could run as a stand-alone service scheduled by the ETL pipelines. Post-PoC with the on-shore team, I provided training materials and mentoring for the off-shore team tasked with the post-Day 1 operations and maintenance.● Tools used in delivery: Python, SQLAlchemy, SOLID Software Architecture, GitLab CI, Nexus Repository Manager, MongoDB

As a director position with incentive benefits, I supplied oversight and delivery aid to provide Gathi/Infostretch/Apexon's world-class consulting and programming deliverables in the finance and medical sectors. During my engagement, all quarterly and annual revenue incentive targets were met for executive bonuses.

● Migrate auction clearinghouse software: My team supplied migration from a C#/Win32 legacy application requiring manual execution to a container-based project hosted on a hybrid Rancher Kubernetes cluster. The code for the infrastructure (Ansible, OCI containers, AWS RDS Posgres), data definitions (SQL DDL), and execution (Java) was hosted in GitHub and deployed by Harness into the Rancher system.

● New bank customer revenue application: Supplied consulting on architecture and mentoring for the team to develop an AWS Step Function TypeScript application, using the codat.io OpenAPI to federate corporate borrowers' operating account transaction ledgers to a data lake in S3 designed to feed ML processes that figure out daily remittance against obligations. Designed the PoC and API client library, and after training the first developer team, supplied education and onboarding to an offshore team to continue the project. The project supported isolated authentication and state for each borrower account link and used AWS EventBus to manage the external webhooks as a service via SQS, SNS, and Step Function subscribers.● Unfortunately, due to technical barriers and business events, this project did not promote production.● Tools used in delivery: AWS Step Functions, TypeScript, OpenAPI specifications and client libraries, AWS S3, AWS SNS, AWS SQS, Harness, SonarQube.

● Migrate auction clearinghouse software: My team supplied migration from a C#/Win32 legacy application requiring manual execution to a container-based project hosted on a hybrid Rancher Kubernetes cluster. The code for the infrastructure (Ansible, OCI containers, AWS RDS Posgres), data definitions (SQL DDL), and execution (Java) was hosted in GitHub and deployed by Harness into the Rancher system.

● Designed, coded and delivered a Python Data Quality Engine using enterprise approved OpenShift containers, python (3.6 only), and JFrog XRay managed libraries.● Built the first Talon Batch deployment into production in the enterprise, using ControlM to execute Python Jobs on the OpenShift cluster with access to both legacy SQL (Netezza) and future-state SQL (Snowflake)● Trained a 4-member team for Day 2+ management and maintenance of the engine and handed over the codebase.● Consulted for the Domino Data Lab Python Notebook cluster migration to AWS, with a desired future state of replacement with AWS SageMaker● Tools used in delivery: Python, GitLab CI, JFrog XRay and Artifactory PyPi and OCI, OpenShift 3 and 4, Control-M, IBM Netezza, Snowflake, AWS, Domino Data Lab (Jupyter notebooks), AWS SageMaker

WAITR provided over 110 cities with employee-drivers to deliver food on demand, using a phone app and dedicated iPads in restaurants.● Managed all AWS services supporting the enterprise, keeping critical systems and supporting 24x7 operations.● Migrated EC2 based services to Spot Instances and dynamic AutoScaling groups. Migrated instance-based MySQL and PostgreSQL databases to AWS RDS Aurora, reducing monthly spend by over 60% and decreasing latency.● Created CI/CD pipelines with CircleCI, AWS CodeBuild and CodePipeline to continuously build GitHub repos into artifacts ready for QA/UAC testing and eventual deployment. This included loading OpenAPI specifications into AWS API Gateway and updating to migrate production traffic to the new deployments. This reduced the 2+ day deployment ritual to less than 3 hours.● Refactored monolithic application into AWS ECR hosted containers deployed into AWS Elastic Beanstalk, using AMD EC2 instances. This reduced outage on “stampede days” and decreased overall monthly costs of the primary user (revenue generating) application by over 20%.● Centralized all logging into AWS CloudWatch Logs and AWS Kinesis Firehose to create a Parquet data lake partitioned by hour, saving storage costs and improving the Analytic Team’s data visibility from a 1-day delay to less than 2 hours.● Integrated with external systems using Lambda Step Functions and AWS EventBus API endpoints. An HR integration reduced the latency between job posting and candidate registration from a semi-daily FTP load to a real-time update, while reducing labor and processing costs.● Negotiated an AWS Enterprise Agreement for custom pricing (requires a minimum of $100,000 MRR to AWS) for the company AWS accounts, further reducing variability in costs and supplying added volume savings on AWS services critical to the business.

As an AWS Premier Tier Partner, Trek10 is in the top 10% of all AWS partners globally. As a Senior Cloud Engineer on the Professional Services team, I was involved with:● Engagements with AWS Professional Services, where AWS supplied on-site technical resources that coordinated under a Trek10 engagement contract to deliver our world-class serverless solutions. AWS ProServ would manage the project, introduce the customer to Trek10, and our teams would drive delivery (Python, legacy systems, AWS).● Engineering for several AWS QuickStart templates published by AWS to provide best-practice and PoC deployments of several services, including AWS CloudFormation StackSet integration with Jenkins CI/CD, AWS CodePipeline solution for multi-party approval flows, and several Lambda serverless event triggered examples. Coordinated with stakeholders on multiple continents had to deliver these projects.● Consulted on-site with customer engineering teams and subject matter experts to overcome technical resistance to cloud adoption. Provided mentoring, training and oversight on solutions while ensuring the customer's access to relevant technical and educational material published by AWS (some covered under NDA with AWS)

● Was the Hesburgh Libraries' Lead Architect for the ND CloudFirst Initiative, a 3-year project to lift and shift 80% of all university computing resources to AWS. The Hesburgh Libraries was the capstone department in the project closeout presentation by the CIO.● Mentored and guided the existing Enterprise Systems Unit administrators into using Infrastructure as Code (Ansible and CloudFormation) for migration and deployment of compute and storage assets into AWS. The team went from 0% cloud experience to multiple AWS certifications per team member (>100% certification)● Supplied GitHub and GitLab CI blueprints and supported the library engineering teams, supporting continual releases into AWS for web and data needs.● Hesburgh Libraries' lead representative for the Cloud Center of Excellence Congress created by the Office of Information Technology's CloudFirst Team. Worked with OIT members and representatives from all campus departments to supply architecture and tooling support for AWS technology initiatives. Aided with first use case documentation through approval via multiple Architecture Board approval processes.● Provided the first campus integration with Talend for integration (AWS PrivateLink and SQS) with the campus ID badging system, achieving the project goal of student ID badges being activated in the library system while significantly improving on the SLO of within 10 minutes of production.● Used Apache NiFi deployed in AWS to integrate with Indiana's ILLiad borrowing system (ILLiad has no API, only a web based human UI), allowing instant automated synchronizing between the legacy library catalog system and external Indiana institution's assets, saving almost a week a month in labor costs ongoing.

● Architected and implemented IBM Cognos TM1 consolidation and reporting cubes, allowing the Finance Team to roll up 6 different operating company charts of accounts and financial data into a single EBITDA report. This system also allowed budget allocations to be projected down into the constituent company's charts of accounts. Each company had different implementation of fiscal calendars, accounting periods, and payroll processes. Extended this system for KPI reporting and feeding as a source of truth for Microsoft SSAS cubes, with drilldown reporting capabilities.● Architected, deployed, and supported a cloud-based Alfresco Document Management 2-tier HA cluster (migrated from an existing single 32-bit instance with no user downtime). The system supported 10,000+ document images ingested, classified, and indexed per day for multiple years. This system provided the HR and Finance teams with indexed and referenceable documents and workflows to support discussion, delivery, and archiving.● Architected, implemented and supported a GitOps ETL process using Pentaho Data Integration and Microsoft SSIS to extract and transform data into an Anchor Model Master Data Warehouse (6th Normal Form).

Consultant for several long term projects, including: ♦ Design data center, and configure servers (AAA, email, news, http, ftp, etc) for several regional ISPs, from a few analog lines to several dozen PRIs ♦ Consulting for several industries such as Medical, Retail, Real Estate, Insurance, Utilities, and Legal offices. Services included line of business database, email, file and print server maintenance, network wide upgrades, backup provisioning and testing. Provisioning included network backbone and campus wiring, digital PRI tandem Class 5 to key system (Option 11e, Mitel, Executone, etc.) service allowing protocol mapping, call routing management, voicemail to email attachments, email to paging gateway (TAP/FLEX) service provided by Intel 486 servers running Asterisk on Linux. ♦ OpenVPN mesh overlay network for the region using OpenWRT, across all service providers for service performance recording (using MRTG), troubleshooting, technician dispatch, and failover to cellular service if needed. Features were managed as individual cost options and prepaid service discounts. ♦ Industrial quality DNS subscription service replicated across the region. This service was capable of scheduling DNS updates to the second (both server side, and client side TTLs), to allow completely transparent global updates to the zones managed. This also included access to the TLD servers via API for direct registration and updates of domain registration records. (using djbdns and OpenSRS) ♦ A backup SMTP service capable of backup MX spooling and resubmitting hundreds of thousands of emails using ESMTP pipelining, allowing network and IP service outages to be transparent to business correspondence and provide minimum latency for delivery interruptions. ♦ Migrating satellite weather downlink from SCO Xenix/286 to a Linux machine, using binary emulation and custom serial hardware.

Brought up a digital-first, TCP/IP native Internet Service Provider using "Official" Red Hat LiNUX 3.0.3 (Picasso) and a commercial Red Hat Support Network subscription (eventually became RedHat Satellite). This OS was installed on Intel and DEC Alpha machines in the NOC for support.Founded several local users groups (e.g., Lafix - the Lafayette Linux Users Group) to provide training and support in Linux and internet tools. The training in HTML, PHP/fi, Apache, FTP, email, faxing helped create several "Web Design" and consulting businesses, some of which are still in operation into the 2020s (e.g., Firefly Marketing Solutions, Rader Solutions).

Brought the first take-all-comers ISP to the 318 area code, from an existing multi-line Novell 3.11 based Bulletin Board System (BBS) a FIDOnet Net 3803 (Terrabyte BBS) peer, UUCP relay, and BBS Door terminal server for telnet, gopher, and FTP. After a migration to Novell 4 and addition of a SLIP NLM, the system became a full TCP/IP gateway for subscribed dial-up users.The local RBOC partnered to provision the first commercial Internet broadband connection (channelized DS1, aka T1) in Lafayette with Sprint Communications. This became Lafayette's first non-academic, locally hosted Internet Service Provider, and the first provider with a local 318 (eventually 337) dial-up number for Lafayette residents and businesses.Once Cisco 250x equipment became available for terminal servers, the system migrated from individual US Robotics modems to a US Robotics Modem Rack (supplied by Cisco), and eventually SIMM based DSP cards. This service became a SLIP/PPP native TCP/IP dial-up Internet Service Provider. A Slackware/SLS 1.00 (from 24 3.5" floppies, provided by Patrick Volkerding personally) with a 3COM 3c501 thinnet ethernet adapter was provisioned for TACACS+, httpd (first Cern httpd, then the patched one, aka Apache httpd) for ~home page service (eventually adding PHP/fi), NNTP, SMTP, POP3, bind (DNS), telnet and finger services.When Livingston digital terminal servers became available, a migration to PRI signaling was completed for a completely digital path from user to the IP backbone. As Livingston RADIUS was on the IETF Standards Track (eventually RFC2138) and had better accounting support than TACACS+, the AAA service was migrated. This provided ISDN BRI 2B+D service to capable clients, and supported MPPP across terminal servers, for channel bonding (aka Multilink).