David vonThenen
Expert AI Engineer and Educator Crafting Machine Learning Systems and Scalable Architectures
E-mail: davidvonthenen [ at ] gmail.com / GitHub: dvonthenen
Summary
Proven technical leader with managerial experience for teams large and small, leveraging decades of expertise in Artificial Intelligence, Machine Learning, cloud-native architectures, storage architecture, distributed network applications, solutions engineering, and developer advocacy to drive adoption and foster strong online presences.
Work Experience
Deepgram (Oct 2023 – Present) – Staff Engineer / Developer Advocate
Staff Software Engineer tasked with forward-looking initiatives and guiding strategic direction to increase platform adoption and sales growth. This encompasses diverse duties, from crafting SDKs, facilitating software integrations and solutions, offering technical insights for the developer ecosystem, writing blog articles and instructional content, and helping with the company’s marketing campaigns.
Implemented major refactors for 2 SDKs (Python and .NET) and introduced the Go SDK as a new offering in the Deepgram catalog. This involved planning a sustainable and maintainable path forward for future features on the Deepgram platform. There was a direct correlation between SDK growth, adoption, and stickiness. Growth results from my direct efforts in 6 months (Jan to June 2024):
o Python 10x growth (10M to 118M calls) – https://github.com/deepgram/deepgram-python-sdk
o .NET 5x growth (1M to 5M calls) – https://github.com/deepgram/deepgram-dotnet-sdk
o Go (Net new 5M calls) – https://github.com/deepgram/deepgram-go-sdk
Worked closely with customers to build out advanced conversational AI solutions. This has resulted in numerous contracts being closed from these efforts. I have been a part of creating solutions including but not limited to the following:
o Creating custom Retrieval-Augmented Generation (RAG) solutions leveraging Large Language Models (LLMs) using vector and graph databases.
o Training custom Natural Language Processing (NLP) models for deriving conversation context based on customer datasets
o Rapidly prototyping solutions tailored to specific customer needs.
o Architected solutions for on-prem, cloud and hybrid deployments in conjunction with customer requirements, infrastructure, and budgets.
Accountable for expanding digital footprint, actively engaging in discussions within the conversational AI and generative AI sectors. Delivering presentations at numerous conferences focused on conversation intelligence and Generative AI, such as:
o (PENDING) Workshop: Building Multiple Natural Language Processing Models to Work In Concert Together – ODSC West 2024, Oct 2024
o (PENDING) Keynote: Training Machine Learning Classification Models for Creating Real-Time Data Points of Medical Conditions – Real Time Communications Conf – IEEE International, Oct 2024
o (PENDING) Building Multiple Natural Language Processing Models to Work In Concert Together – Real Time Communications Conf & Expo – IEEE International, Oct 2024
o Empower Your AI Journey: Hands-on Machine Learning with fastai for Graphics and NLP – AI Dev Summit 2024
o Voice-Activated AI Collaborators: A Hands-On Guide Using LLMs in IoT & Edge Devices – SCaLE 21x, March 2024
o Enhancing Real-Time WebRTC Conversation Understanding Using ChatGPT – Real-Time Communications Conference & Expo, Oct 2023
o Edge Devices as Interactive Personal Assistants: Unleashing the Power of Generative AI Agents – Real Time Communications Conference & Expo, Oct 2023
Created blog content covering various topics, encompassing technical integrations, project launches, promotional initiatives, and insightful industry analyses.
o Lone Wolf vs Community: The Benefits of Open Source Software
o The Code Revolution: Python SDK v3.0 Now Available
o Announcing Official Go Support: v1.0 Released for General Availability
Symbl.ai (Aug 2022 – Aug 2023) – Principal Developer Advocate
Hired to create the developer advocacy program to provide a blueprint for the company. This included creating SDKs, developing software integrations, providing technical expertise in customer sales meetings, authoring blog posts and training material, and contributing to the company’s social media/marketing campaigns.
Lead Software Engineer on the Enterprise Conversation Application project. Allows customers and individuals looking to leverage a framework to jump-start creating their conversation understanding applications. The project implements a plugin infrastructure functionality that allows creating plugins to customize the framework to their specific use cases. Blog posts describing the project are below:
o Everything to Know About Enterprise Reference Implementation for Conversation Aggregation
o Introducing an Easy Button for Creating a Pluggable Enterprise Conversation Application
Lead Software Engineer on the Go SDK. Founded the project and has had 12 releases for public consumption. The SDK has been instrumental in landing numerous deals. The initial account landed was valued at 60k+. This would later turn into continued revenue growth.
Implemented first Symbl.ai LangChain integration for Nebula, Symbl.ai’s conversation LLM. The LangChain integration allows other users of LangChain’s large ecosystem to build projects and products by enabling Generative AI applications to use Nebula to deconstruct conversations used in conversations, chatbots, etc.
Responsible for growing online presence, including contributing to the conversation and generative AI space. Presenting at a variety of conversation intelligence and Generative AI conferences, such as:
o Streamlining Communication Workflows – Nexus x TPF GenAI Rush 2023
o Enabling Untapped Use Cases in ML/AI – API World 2022
o Voice 2022 Hackathon – Voice & AI 2022
Authored 20+ blog posts on various themes, including technical integration, project releases, marketing campaigns, and industry thought-provoking pieces.
o Know the Security and Privacy Implications of Conversation Data Retention
o Understand Your Business Better with ChatGPT and Symbl.ai Platform Integration
o Extract Insights using Symbl.ai’s Generative AI for Recall.ai Meetings
o Databases and Persistent Storage for Conversation Data
Created the Symbl.ai Video Training Series (instructional YouTube videos) that walked developers through features on the Symbl.ai Platform. The videos targeted technical and non-technical users. Examples of some of those videos:
o Symbl.ai Platform Training Series: Topics
o Symbl.ai Platform Training Series: Redaction
o Symbl.ai Platform Training Series: Entity Detection
VMware (Oct 2010 – Aug 2022)
Modern Applications Platform Business Unit (Feb 2018 to Aug 2022) – Staff Software Engineer
Cloud Native Engineer and Developer Advocate, who provided the initial foundation for VMware’s Kubernetes strategy. Architected and implemented multiple critical Kubernetes services that enabled VMware to operate as a Kubernetes cloud provider. This required writing an enormous amount of software development and evangelizing initiatives in the Kubernetes ecosystem.
Founding Member for developing and launching Tanzu Community Edition (TCE), an open source Kubernetes distribution that provides unique integrations with the Carvel toolset. TCE provided a fully integrated Kubernetes experience to create management and workload clusters and deploy application packages.
o Created the initial implementation for a pluggable application deployment package mechanism
o Created the CI/CD automation to generate the downloadable artifacts for the TCE distribution.
Responsible for growing my online presence, which includes contributing Kubernetes and Cloud Native Computing Foundation (CNCF) related blogs, creating content for my personal blog, engaging with users in Kubernetes Slack, and presenting numerous conference sessions below:
o Learnings From Creating CI/CD Pipelines for Open Source Kubernetes Projects – Cloud Native Rejekts Europe 2022
o Provider vSphere: All Things vSphere Working Group – KubeCon + CloudNativeCon Europe 2020
o Intro: Kubernetes VMware SIG – KubeCon + CloudNativeCon Europe 2019
o Application Monitoring and Tracing in Kubernetes: Avoiding Microservice Hell! – Open Source Summit Japan 2018
o Application Monitoring and Tracing in Kubernetes: Avoiding Microservice Hell! – SCaLE 16x 2018, Los Angeles
Contributing member and project maintainer for Cloud Provider vSphere, an open source Kubernetes project enabling Kubernetes to be infrastructure aware in on-premise vSphere and VMware Cloud on AWS (VMC) environments. The project provides capabilities for load balancing, computing resource scheduling, region/zone support, multi-tenancy, etc.
Designed and implemented the first community public offerings for the vSphere Container Storage Interface (CSI) Driver for Kubernetes. This open source project provides the capability to provide various types of storage and schedule container workloads that require data persistence. The CSI Driver provided file and block storage that implemented the Kubernetes best CI/CD practices through validation (lint, vet, etc.), unit tests, and end-to-end testing via Prow.
Dell/EMC – Emerging Technologies Division (Oct 2015 – Feb 2018) – Staff Software Engineer
Transferred into Dell/EMC’s Emerging Technology Division where I took on Developer Advocate role with the Dell/EMC {code} team. I was tasked to create open source projects and contributions to the Kubernetes and Docker ecosystems that aligned with Dell/EMC’s software initiative. The {code} team and projects were transferred to VMware (above) to increase acceleration/adoption.
Responsible for growing my online presence, contributing to the team’s official {code} blog, creating personal blog content, engaging users on the community slack channel, and presenting numerous conference sessions below:
o Robust Applications in Mesos using External Storage – MesosCon EU 2017, Prague
o How Container Schedulers and Software-based Storage will Change the Cloud – ApacheCon US 2017, Miami
o Demos, Demos, Demos! Containers and {code} – Dell EMC World 2017, Las Vegas
o Managing ScaleIO As Software on Mesos – Dell EMC World 2017, Las Vegas
o How Container Schedulers and Software Defined Storage will Change the Cloud – SCaLE 15x 2017, Los Angeles
o Open Source Storage Summit – ContainerCon EU 2016, Berlin
o Software Defined Storage and Container Schedulers – ContainerCon EU 2016, Berlin
o Deep Dive with Mesos & Persistent Storage for Applications – EMC World 2016, Las Vegas
o Introduction to Mesos and Mesosphere – EMC World 2016, Las Vegas
Kubernetes Contributor. Worked with SIG Storage to provide storage-related contributions. Helped foster the creation of the Container Storage Interface (CSI) Working Group to establish a generic storage interface for orchestrating storage operations for on-prem and cloud providers. This would, in turn, establish the foundation of the CSI Specification during my time at VMware.
Kubernetes Helm and Charts Contributor.
o Provided the initial implementation of the Jaeger Helm chart, which provides an easy button to deploy and manage the lifecycle of a production Jaeger instance.
o Contributed functionality to the Istio chart to support Jaeger tracing for mesh networks
o Adding dev/test capabilities to the ElasticSearch chart.
Developer on the mesos-module-dvdi open source project, which provides a pluggable Docker Volume Driver Interface to manage storage volumes for AWS EBS, AWS EFS, GCP persistent disks, Azure-managed disks, etc. This project was eventually consumed by the Apache Mesos project.
Architected and implemented the ScaleIO Framework for Apache Mesos, enabling storage protection pools and resource access on Mesos. Supported provisioning from bare metal to cloud-based resources with cloud-aware monitoring and remediation on AWS and GCE.
Dell/EMC – Backup/Recovery Solutions (Oct 2010 – Oct 2010) – Sr. Principal Software Engineer
Lead Software Engineer turned team manager for 20+ on Backup/Recovery solutions specializing in VMware Virtualization backup. Grew the business unit’s revenue from $880 Million to $2.18 Billion, and 80% of that revenue increase was directly attributed to our development of virtualization backup.
Architect and developer for EMC Avamar’s VMware Virtual Machine Backup/Restore solutions.
o Designed and implemented critical features for image-based backups/restores, including Change Block-based synthetic full backups/restores and application quiesced backups for SQL, Exchange, and SharePoint.
o Drove implementing and adopting file/folder-level recovery on image backups. This required block and device storage internals to expose disk and volume mounting to support numerous Linux and Windows filesystems.
Developer for EMC Avamar’s vCloud Protector (vCP), a backup and recovery solution for VMware’s vCloud Director (vCD) and vCloud Air. This provided backup/recovery at cloud scale. Enabled replication of on-premise backups to vCloud Air and seamless recovery to/from the cloud.
Established a robust Continuous Integration (CI) environment and software development process that isolated release engineering from integration issues, introduced automated unit tests, code reviews, and regression testing on developer branches, significantly reducing defects by 60% between product cycles.
QLogic Corporation (Jan 2006 – Oct 2010) Staff Software Engineer
Developed and optimized remote management applications and drivers for Fibre Channel (FC), Fibre Channel over Ethernet, and iSCSI storage adapters, including ASIC firmware programming for manufacturing. Every 4GB and 8GB storage adapter shipped was programmed and managed by software I had built.
Unisys Corporation (Sept 2003 – Jan 2006) Software Engineer 4
Developed several products relating to cluster management, proactive, efficient failover of applications, and creating highly available platforms.
Intel Corporation (April 2003 – Sept 2003, 6Month Contract) BIOS Validation Engineer
Developed test cases on production releases of system BIOS for several Multi-processor motherboards and microcode used on Intel Xeon processors.
Conexant Systems/Mindspeed Technologies (Feb 2001 – June 2002) Software Engineer
Developed multiple automation applications used by engineers in the semiconductor design process. The chips I had direct impact on at Mindspeed was low latency, high throughput internet switching chips.
Bachelor of Science in Computer Science, University of California, Irvine
Specialization(s): Networks/Distributed Systems, Embedded Systems
Bachelor of Science in Chemistry, University of California, Irvine
Specialization(s): Organic Chemistry
Artificial Intelligence / Machine Learning:
PyTorch, Keras, Transformers, Generative AI, NVIDIA CUDA, Pandas, Large Language Models (LLMs), ChatGPT, OpenAI APIs, LLaMA by Meta, Autonomous Generative AI Agents, Pinecone, Chroma, LangChain (Python, Go)
Programming Languages/Technologies:
Go(lang), Python, C/C++, Java, JavaScript, Cloud Native Architectures, open source software development, Cross-platform development (Linux, Windows, MacOS), gRPC, REST APIs, SQL (MariaDB, PostgreSQL, MSSQL), NoSQL (Elastic Search, Cassandra), Graph DBs (Neo4j), CI/CD (using Jenkins, Travis, GitHub Actions), Security and Encryption, GitHub/GitLab, GNU Make, Maven, RPM/DEB creation, PowerShell, bash, git, Mercurial, Perforce
Kubernetes, Docker, Prometheus, Jaeger, OpenTelemetry, creating Helm Charts, container Image creation, etcd, AWS Lambda, Amazon EKS, Azure Kubernetes Service (AKS), EC2, S3, CloudFormation templates, IAM, AWS CLI, Go AWS SDK; Google Compute Engine, GKE, Compute Engine APIs in Go
VMware vSphere Environments, Hyper-V, KVM, Xen, Aria/vRealize Suite, Aria Operations for Logs (formerly vRealize Log Insight), VSAN, Virtual Volumes, vCloud Director Environments, HA Clusters, Storage DRS, Fault-Tolerant VMs, vStorage APIs for Data Protection (VADP), vSphere SDK, Virtual Disk Development Kit (VDDK), VMware Mount API, PowerCLI
Developer Advocacy & Evangelism:
Brand Creation, Executing Marketing Campaigns, Managing LinkedIn Company Page and Company Twitter, Creating/Publishing Marketing Videos (for Corporate Presentations, Training, Community Meetings, Office Hours, Demonstrations), Authoring Blog Posts (for messaging and technical content), Hootsuite, SocialPilot, URL and UTM Tracking, Adobe Cloud Suite (using After Effects, Illustrator, Photoshop) for Video Editing, Screenflow, OBS Studio, Asciinema
US9292327B1 – Optimization for incremental backup of VMs
US9389964B1 – Read-write access in a read-only environment
US9026500B1 – Restoring virtual machine data
US9355235B1 – Validating a user of a virtual machine for administrator/root access
US10467034B1 – Performing application specific actions without in-guest credential information