Introducing hotg.dev - open ideas meets open source
Rune is our open source MLOps container for edge computing
The content of this post was posted last week on LinkedIn by our CEO Akshay. Since we have a wide following here on Tinyverse, we thought it would be appropriate to share it here as well.
Our team will also move all bi-weekly updates to monthly updates from the technical team. Bi-weekly updates will be replaced with technical driven posts as well as some guest appearances from our circle of incredible networks. We are excited to bring this new change to Tinyverse. If you would like to guest blog, feel free to drop Holly a line, our Head of Operations & Community Relations at holly@hotg.ai.
I wanted to share our ethos of building HOTG.AI as an open-source product company, and our culture to rally around the community first.
First, a few things to note, that will help understand our thesis:
Pandemic changed the culture of deep tech. You can be anywhere in the world and build things.
The idea of time, timezone, and location is obsolete - in fact, you are giving up easy alpha not being distributed from get go.
The creator and builder economy emerged almost immediately. Building a product without community is like licking the tip of a finger and sticking it into the air.
The future is more decentralized due to talent distribution.
These are the same market forces that benefit us at HOT-G building the MLOps platform for the Edge.
Open source is about building trust. When you run workloads on servers, cloud, and even phones, you are putting a lot of trust in it. One way to scrutinize this trust is by auditing the code that runs on these devices - and open source makes this easy. Open source is more than this - it is a culture of building with a community and rallying around common beliefs. We also think
Some of the best talent work on open-source projects
Open source is already distributed, and setup correctly has every aspect of decentralized decision making possible
For us open source is how we will get to scale. Building container technology which helps solve the MLOps for the edge, along with the surrounding infrastructure of security, reliable builds, and telemetry means we have to scale frameworks (TF, TF-JS, Pytorch ...) across a fragmented world of devices (mobile, web, desktop, sensors, MCUs, etc). Open source is how we become Pareto efficient.
Every engineer in HOT-G is pushing something that contributes to our open-source effort.
You cannot be an open-source company without making it a key investment upfront (money, time, and culture). The challenge that comes with open source is how do we manage our product lines (SAAS and enterprise) while keeping our roadmaps coordinated. In an ideal world they are all the same, but this is not the case. We usually see one or the other having a more recent tip of the tree. However, setup right, we know it starts to compound the value very quickly. For example, our security bits started enterprise first and only then into our OSS SDKs.
We have seen companies like Hashicorp and Databricks (for Spark) do an excellent job with this model.
What does this mean to you?
If you are an engineer, builder, maker, tinkerer, community stewards, or simply curious kind come join us in forging the field of Edge SecOps. We wrote about this in our newsletter -
We have re-launched hotg.dev to fully focus on our community and OSS offering.
Here are a few ways you can already contribute today by visiting hotg.dev:
Give us github stars for our open-source container technology - https://github.com/hotg-ai/rune
Fork or clone it and start running it. You can do all the things we do in our SAAS platform - https://studio.hotg.ai using the same CLI tooling. This is the power of OSS - you don’t have lock-ins - even with us!
Even better - you can add support to another ML framework or a device - hit us up on discord on how to
Join our discord https://discord.gg/gPCNNvRnF4 and chat with us, watch our community interactions, and be part of this new wave
Kartik and I as founders have a lot of plans to bring more collaboration features and build the dev community around this.
Join us and have fun!
#opensource #ai #community #mlops
— Akshay