We've read your feedback, heard your criticism, looked at the Lyrid platform and thought “what can be changed??” I’m excited to announce a new batch of Lyrid feature updates that’ll bring your development and deployment processes from start to finish as fast as possible!
Expect changes and improvements to our deployment, documentation capabilities, cluster API support, and so on. Whereas for the future of Lyrid, look towards the AI integrations, data platform integrations, new support for existing services, and so much more.
Lyrid has been pushing sales efforts for Q1 and Q2 of 2023. And one big feature request from our data center partners like Biznet and Serverpronto, and from our public cloud partner, Oracle, is to support Helmchart deployments. Now, you'll be able to deploy Helmcharts in different configurations on specific clusters dynamically.
TODAY, Helmchart deployments are available for data center partners, limited beta users—send me an email firstname.lastname@example.org if interested— and enterprise Lyrid users! This feature works in tandem with our cluster creation and onboarding features all through a seamless and easy to use UI— aka 3-4 clicks per feature. 👏
A big inspiration to Lyrid’s market positioning is seeing the gap in platform-as-a-service (PaaS) companies like Vercel, Heroku, etc. and infrastructure-as-a-service (IaaS) companies such as OpenStack. In order to bridge this gap, one feature that will 10x ease of use in Lyrid is a one-click-deploy button, popularized by Vercel.
Starting next week, you’ll be able to deploy an application/service from an open GitHub repository directly to Lyrid using our One Click Deploy button. The process is pretty simple, you just need to edit the yml.file to Lyrid configuration and add the button to the public repository and anyone from your user base can automatically deploy in Lyrid after account creation. For more details, check out the next section below!
The floodgates have opened! The rocket has launched! More and more users are coming into the Lyrid platform and one BIG error we made was not updating our documentation enough. But no longer!
What can you expect?
This is our biggest documentation update in over a year, so bear with us. We’ll make incremental changes in the next few weeks, so if there are errors or if you have any specific documentation request let me know! P.S. We’ll be releasing a technical blog on “How to make a custom AI project on Lyrid” soon…
If you’ve run your applications on Lyrid on a production level, an annoying bug that was persistent after our last big update was having to redeploy your application/solution each time there is a revision. But no longer.
We heard you! 🙂 No more manual redeployment. No more hassle. Our backend can handle any application, solution or microservice update you re-submit to Lyrid while keeping current deployment with no downtime. But of course, if you do want to reset the application deployment, we’ll still give you the option.
Now this is big. While a large portion of dynamic cluster onboarding and creation is still handled in our backend, we’re starting to roll out our Kubernetes cluster API support directly to our Lyrid Enterprise clients. For our first version release, we’ll support OpenStack and Oracle such that users can buy and onboard Kubernetes clusters onto the Lyrid platform dynamically through a couple clicks. (insane! Am I right ?!?!)
For our partners, this is exciting news!! Lyrid users can buy Kubernetes clusters directly from data center partners such as Biznet and Eranyacloud in Indonesia or public cloud partners such as Oracle. We’ll be able to expand our local data center support and make big strides in our localization strategy. If you’re interested in partnering with us, and being a part of this — again, I’m your guy email@example.com
Protecting your applications, machine clusters and most importantly, your business within the Lyrid platform are of the utmost importance to Lyrid. This past weekend, you might have noticed an email that prompted you to reset your Lyrid password. We have decided to push this as a regular security update, safeguarding your account as well as refreshing and bolstering the security of our systems.
If you haven’t done so already, please check your inbox for that email, or visit https://app.lyrid.io/forgot to reset your password, we apologize for any inconvenience and thank you for your participation!
December 2022, the month everything changed. Or so they say. In our industry, and worldwide, when OpenAI released ChatGPT to the public, a craze like never before hit the startup scene. It seems like everyone was making their own AI startups. As a platform that bridges cloud infrastructure with the developer experience, Lyrid has to adapt quickly to this changing market. As such, there are big changes coming to the platform for AI startups.
From data ingestion to ETL pipelines to vector databases to GPU machines to train your models … and many more. If you want to be in the waitlist for the beta access for any of these features, hit me up firstname.lastname@example.org or on Twitter @reinhardwilmer.
A key ingredient in a useful data platform is a robust data pipeline. So this is what we came up with. This is an overview of the system diagram that we’re developing inside Lyrid. See below.
These use the best of all Open Source Software data solutions that are well maintained, have very strong community support, and have enterprise backing to be able to ensure the longevity of our solutions. We even use these data pipelines internally as our own internal infrastructure and business analytics.
Now let’s take a deep dive into the architecture of the data platform and walk you through every component that makes Lyrid Epsilon what it is.
By default, Epsilon utilizes Apache Superset as the engine to visualize our charting inside our application. This gives users a native for the application to build drag and drop charts and dashboards without writing SQL queries, the ability to embed charts on other applications using HTML embedded iFrame, along with additional features such as slack & email alerts support.
Additional to Superset, we also expose Presto/Trino endpoint externally because it can seamlessly integrate with many external tools that run many different solutions for different use cases that use data like:
On the data ingestion side, Lyrid supports Rest API and Apache Kafka integrations for data streaming, and handling API calls. The ETL warehousing is pretty interesting, on our beta release, we’ll mainly use Airflow to handle the ETL load, scheduling, and jobs through custom DAGs available. The warehousing itself will be supported using PostgreSQL and MinIO. In later releases or in big data use cases, Epsilon has a Delta Lake and Apache Spark support — reach out to me @reinhardwilmer on Twitter or email me if you need this.
Lyrid’s goal since our inception is to bring technology access to the masses, and one area of LLM that needs disrupting is in the vector embedding space. Right now the current top dog is Pinecone, but even they can’t handle all the vector embedding demands, and thus, we turn to Weaviate as a more viable alternative.
Weaviate is an open source vector DB with semantic search capabilities — pretty much all the functionality you want from a vector DB. With our unique value prop of data center localization and ease of use, we can truly make a unique offering here for AI startups. You can run your very own models + vector DB in your own machines (GPU, RAM, Storage all customizable), the best part is costs scale linearly vs public cloud cost structure that scales exponentially (no bs., no hidden fees). So yes, Weaviate support is coming real soon, join the waitlist here and save some BIG BUCKS $$$.
Slightly non-AI related feature, but due to increasing demand in support for dynamic public cloud Kubernetes cluster provisioning, we will be adding support for Google Cloud Platform (GCP) using their Google Kubernetes Engine (GKE) through Kubernetes Cluster API.
This would be great if you’re in the Google Startup Program and still using your Google credits. Just an fyi, we work with startups using public cloud credits all the time. In fact, Lyrid can provision GKE compatible Kubernetes clusters dynamically using your credits.
So why do it with Lyrid? Simple. Once you’re in Lyrid, you’re not vendor locked — move public clouds, on-prem, or even totally isolated machines anytime. If you’re anticipating a cloud spend hike, this is the right time to consider a move — we’ll make it as easy as pie.
Current timeline: limited beta access will be available by the end of Q2 2023. Further support for other public clouds such as Amazon Elastic Kubernetes Service (EKS) and Microsoft’s Azure Kubernetes Service (AKS) will be developed depending on user demand (so if this is something you want, hit me up—email@example.com)
A necessary gap in our Enterprise support offering so far is improving our notifications, monitoring and logging architecture. As a part of the big update coming up, let me break this down for you:
The notification system feature will provide users with a way to stay informed about the status of app deployments/management, database, and Kubernetes clusters. Users will be able to choose from multiple notification channels and configure their preferences to receive the types of notifications that are most relevant to them.
The monitoring system will provide users with relevant data on their applications, database, deployments, Kubernetes clusters and many more, through our analytics engine Prometheus, a highly scalable open-source Kubernetes monitoring framework. We also have Sentry, an application performance monitoring and error tracking tool that will help us notify our users to handle error and downtime related issues before it actually happens.
Lyrid is building a comprehensive and robust logging architecture system where activity logs at the application, node, pods, and Kubernetes cluster levels are monitored and logged in a secured database/object storage environment. The main benefit of this is we’ll be able to diagnose what, when, where, and how events happen at an application and infrastructure level.
At Lyrid, we’re constantly creating new ways for you to interact with our platform and level up your solutions and applications. From new AI and data integrations, to revolutionizing the way you deploy and develop, this batch of new feature updates is a turning point within the Lyrid platform.
Looking towards the future is something that we do daily, and we couldn’t have done it without your support and enthusiasm for our platform (thank you!). If you’re looking to see these rolling updates in action, book a demo with us here!
99 South Almaden Blvd. Suite 600
San Jose, CA
Jl. Pluit Indah 168B-G, Pluit Penjaringan,
Jakarta Utara, DKI Jakarta