Robert Vandervoort

This project I created for demonstrating the use of OpenTelemetry line protocol over http for sending high precision data from low power devices. The project also includes tracing capabilities for timing of stages of device actions such as connecting to wireless networks, establishing time through NTP, and pulling and sending data from sensors connected via I2C serial bus. I wrote a custom OTel library based on OTel standards that is not a derivative work of existing libraries. This taught me a… Show more

This project I created for demonstrating the use of OpenTelemetry line protocol over http for sending high precision data from low power devices. The project also includes tracing capabilities for timing of stages of device actions such as connecting to wireless networks, establishing time through NTP, and pulling and sending data from sensors connected via I2C serial bus. I wrote a custom OTel library based on OTel standards that is not a derivative work of existing libraries. This taught me a LOT even though it may have been unnecessary. The library allows users to take advantage of OTel in very low powered, resource constrained devices that are not typically deployed in commercial IoT use cases, providing for inexpensive,, expendable monitoring in high risk areas.

This project demonstrates the feasibility and flexibility of this mix of technologies in a market segment that has historically been a “walled garden” with big players operating in Scada and ControlNet networks. While I doubt it will replace or even disrupt that current market anytime soon, it provides an alternative for observability practitioners to bring environmental and other data into their observability platforms and take advantage of automated correlation present in many of these platforms. The sky’s the limit as they say.

https://github.com/robertvandervoort/iototeldemo Show less

There are tons of these out there. Mine may not be the most fully customizable regarding looks (though I’m slowly improving that). What it excels at is bulk creation. My wife asked me if there was a way to create a hundred QR codes easily from a database / spreadsheet. I said you can do this pretty easily with code since there don’t seem to be any solutions that did it this way. I turned this app around the same day, saving her a lot of tedious work. Her coworkers asked for more QR types like… Show more

There are tons of these out there. Mine may not be the most fully customizable regarding looks (though I’m slowly improving that). What it excels at is bulk creation. My wife asked me if there was a way to create a hundred QR codes easily from a database / spreadsheet. I said you can do this pretty easily with code since there don’t seem to be any solutions that did it this way. I turned this app around the same day, saving her a lot of tedious work. Her coworkers asked for more QR types like email addresses, locations, and even styling option which I added the same day. I launched the app on a private web server and it has received regular weekly usage from colleagues at cisco and at her office. I initially wrote it in python but she was not able to use it on her laptop at work due to security. The next day I rewrote it in Typescript / React and put it up on a scalable web platform. The user experiences are a bit different. I favor the React version over the python version and as such that’s what will get the new features in the future.

Python version: https://github.com/robertvandervoort/QrCodeGenerator-React
React version: https://github.com/robertvandervoort/QrCodeGenerator-React Show less

This was born of the desire to automate note taking in calls where recording is not allowed. The problem with text to speech is that movie script like outputs are not available. Speech to text is easily stymied by multiple speakers and interruptions. This project uses AI voice models to detect speakers and speaker changes in the conversation, coupled with another model for the speech to text and translation. The solution uses all local models on multiple compute platforms and is quick and… Show more

This was born of the desire to automate note taking in calls where recording is not allowed. The problem with text to speech is that movie script like outputs are not available. Speech to text is easily stymied by multiple speakers and interruptions. This project uses AI voice models to detect speakers and speaker changes in the conversation, coupled with another model for the speech to text and translation. The solution uses all local models on multiple compute platforms and is quick and fairly efficient. Since i am taking advantage of the Seamless M4T v2 model from Meta, it can also do speech to speec. This release is pending quality testing but the aolution is available now without speech to speech at https://github.com/robertvandervoort/ai-dictation

It is my intent to use SeamlessExpressive to mimic speaker tone and cadence and then eventually provide
Options for RVC. Running local models provide security but also add to the complexity of the solution as third party (paid) offerings have the capability today but are not performing speaker diarization. Show less

A tool I wrote in Python to extract, transform and enrich data coming from AppDynamics controllers. This is helping customers and colleagues better understand area for configuration improvements and will eventually contain AI driven auto configuration. This configuration fine tunes instrumentation to best catch anomalies, reducing alert noise.

Implementation of IDERA SQL Server Tools @ Turkish textile giant LC Waikiki.

Responding to RFPs isn’t fun. Let Gemini do it. I slapped this Google apps script together real quick for Google sheets you need a Gemini API key and it will use the vast knowledge of the internet mixed with some hallucinations to take a first pass at those pesky RFPs. Or really any spreadsheet full of questions! You edit the prompt and tell it what columns to use and fire away.

Hint, do review the answers! You’ll be surprised how close it gets and how much work is saved… Show more

Responding to RFPs isn’t fun. Let Gemini do it. I slapped this Google apps script together real quick for Google sheets you need a Gemini API key and it will use the vast knowledge of the internet mixed with some hallucinations to take a first pass at those pesky RFPs. Or really any spreadsheet full of questions! You edit the prompt and tell it what columns to use and fire away.

Hint, do review the answers! You’ll be surprised how close it gets and how much work is saved.

https://github.com/robertvandervoort/autorfp-Gemini Show less