evahteevGuru Idea Flow
How do you approach ideas, and how much do you stick to them once you start following them? When do you pivot to a new direction, turning the flow of previously shiny ideas in a new direction or wrapping things up under a new umbrella when they outgrow the old one? These are questions that often challenge Web2 startup founders and, even more so, Web3. FOMO, volatility, and trends are constantly shifting—bull cycles, winters, and who knows where we are now. We started as a Multi-Chain/Multi-De...
evahteevIn Search of an Infrastructure Holy Grail for DexGuru: Mastering Cost-Effectiveness and Performance
TL;DRDexGuru's transition from cloud-based infrastructure to an on-premises setup marks a strategic shift towards unparalleled cost-effectiveness and control. This move not only aligns with our commitment to full data supply chain control but also enables us to deliver custom-tailored, high-performance solutions, especially for data-intensive applications like OLAP databases.Unrivaled Cost-EffectivenessWe started our MVP in AWS (Got to $170k a month check), then decided to move towards H...
evahteevInscriptions EVM Madness
As we are running our own EVM infrastructure ourselves and not relying on any third party services, we have to deal with all kind of new trendy things allowing bots/users to hit new bottom of Geth performance against hardware. In December it was Inscriptions. For whatever reason someone decided that moving BRC-20 on EVM chains is a good idea, which can be cheaply implemented submitting transactions with nothing but data in Input fieldinscriptions rawParsed easily with Clickhouse JSONExtractSt...
CTO & Co-Founder of dex.guru, block explorer https://b2b.dex.guru/explorer and data warehouse https://warehouse.dex.guru/
evahteevGuru Idea Flow
How do you approach ideas, and how much do you stick to them once you start following them? When do you pivot to a new direction, turning the flow of previously shiny ideas in a new direction or wrapping things up under a new umbrella when they outgrow the old one? These are questions that often challenge Web2 startup founders and, even more so, Web3. FOMO, volatility, and trends are constantly shifting—bull cycles, winters, and who knows where we are now. We started as a Multi-Chain/Multi-De...
evahteevIn Search of an Infrastructure Holy Grail for DexGuru: Mastering Cost-Effectiveness and Performance
TL;DRDexGuru's transition from cloud-based infrastructure to an on-premises setup marks a strategic shift towards unparalleled cost-effectiveness and control. This move not only aligns with our commitment to full data supply chain control but also enables us to deliver custom-tailored, high-performance solutions, especially for data-intensive applications like OLAP databases.Unrivaled Cost-EffectivenessWe started our MVP in AWS (Got to $170k a month check), then decided to move towards H...
evahteevInscriptions EVM Madness
As we are running our own EVM infrastructure ourselves and not relying on any third party services, we have to deal with all kind of new trendy things allowing bots/users to hit new bottom of Geth performance against hardware. In December it was Inscriptions. For whatever reason someone decided that moving BRC-20 on EVM chains is a good idea, which can be cheaply implemented submitting transactions with nothing but data in Input fieldinscriptions rawParsed easily with Clickhouse JSONExtractSt...
CTO & Co-Founder of dex.guru, block explorer https://b2b.dex.guru/explorer and data warehouse https://warehouse.dex.guru/
Subscribe to evahteev
Subscribe to evahteev
<100 subscribers
<100 subscribers
Share Dialog
Share Dialog
The underlying infrastructure layer is crucial for the future success of any AI and blockchain integration, and Guru Network is well-positioned to lead this transformation.By transitioning from cloud-based infrastructure to an on-premises, censorship-resistant, and decentralized network, we have laid the foundation for a robust Infrastructure-as-a-Service (IaaS) model. This model powers individual agents and orchestration nodes across the Guru Network, supporting the utility of $GURU and enabling a dynamic and thriving ecosystem. Our expertise and innovative infrastructure solutions powers up Guru Network as a provider of Multi-Chain AI/Web3 orchestration.
Video Version:
Our infrastructure journey began with DexGuru, initially utilizing AWS for its scalability but incurring significant costs of up to $170k per month. Seeking greater cost-effectiveness and control, we transitioned to Hetzner Cloud/Dedicated until restrictions as a crypto company forced us to build our own cloud completely at some point. This move marked a strategic shift towards unparalleled cost-effectiveness and performance. More on it: https://mirror.xyz/evahteev.eth/POoVA0GiX9XcaQXGwQekS7o57pWwKARYzEW4uofwX2E
Our meticulously planned on-premises infrastructure includes high-performance computing (HPC), advanced storage solutions, and sophisticated virtualization techniques. Key components include:
Kubernetes (K8s): For managing containerized applications.
Elasticsearch: For powerful search and analytics capabilities.
Clickhouse: An efficient column-oriented DBMS.
Redis: Used for caching and as a message broker.
Blockchain Nodes: The backbone of our data pipeline.
This setup provides predictable expenses, complete control over resources, enhanced performance, and improved scalability, particularly for data-intensive applications like OLAP databases.
Guru infrastructure: https://github.com/dex-guru/guru-infrastructure (We started from open-sourcing virtualization first, more coming)
The economy is quite simple here, and with AI Models/GPU requirements it becomes even better, for example to run proposed chatbots:
https://x.com/xgurunetwork/status/1805696565786362135
We widely used Dell servers already, and let’s go through an example here with GPU models running.
During normal hours compute happens there(paid off within 3 months compared to cloud costs for the same compute). But we can redistribute and auto provision new ones on GCP in case of increased traffic.
Same time it has redundancy and scaling of the Cloud in case of Maintenance/Issues or Increased traffic. That’s the core of the Guru Hybrid infrastructure solution. That also allows for future decentralization where node runners would provide distributed compute using the Guru Infrastructure framework to run it in Clouds or On-Premises and earn $GURU. As the first step we, ourselves operating now in multiple environments and releasing the automation scripts and nodes so anyone would be able to do it in future.
And we are not the ones there. I had discussions with many RPC and Data providers operating with blockchain data over the years. Most of them having own on-premises compute and storage centers, which became regular practice within the space. Additional motivation there was crypto winter when everyone were cutting costs and working on cashflow.
Some Web2 projects are also following the same pattern, like the most loud one:
What future awaits us? When cloud providers:
And we answer:
https://mirror.xyz/evahteev.eth/POoVA0GiX9XcaQXGwQekS7o57pWwKARYzEW4uofwX2E
Now with Guru Network ideas and flow we are positioned for more on that front, decentralized and self incentivized
At the heart of Guru Network's infrastructure are Individual Agents—decentralized, autonomous computational entities that execute specific AI-driven tasks. These agents operate as part of our blockchain business process automation (BBPA) engines. By running open-sourced AI models, they enable a rich ecosystem where products and features can seamlessly share context and threads, enhancing the capabilities of our chatbots, image generation tools, and other AI applications. More on Architecture in light paper: https://gurunetwork.ai/assets/img/litepapper_dexguru_network.pdf
The Flow Orchestrator is a critical component, managing the deployment and operation of BBPA engines and AI processes. It provides a low-code development environment that simplifies the creation and management of complex workflows. This orchestration enhances the efficiency of blockchain applications and ensures that AI models can operate effectively across multiple chains and environments.
Our decentralized infrastructure supports a dynamic IaaS model, empowering developers and businesses to leverage our tools to build innovative applications. By publishing our infrastructure as open-source, we foster a collaborative environment where participants can contribute to and benefit from the advancements in AI and blockchain technology.
The $GURU token underpins the economic operations of the Guru Network, serving as a fundamental element of our IaaS model. It incentivizes node runners, contributors, and users, facilitating a thriving ecosystem. Each transaction and process within the network is compensated in $GURU, ensuring a stable and predictable economic environment.
But with all that motivational speech, we should say CLOUD PROVIDERS are crucial for interoperability, CDN level solutions, as well as try fast - deploy wisely. In our Hybrid infrastructure solution they just no occupying every level. Our partnership with Google Cloud has been useful. The GCP program for Web3 startups allowed us to experiment with different combinations of GPU/LLM Models servers and optimize our data distribution layer for performance. As we call it “Hybrid solution”: the compute layer(costly and heavy one) stays within our On-Premises DC, while APIs and FE for B2B clients (Block Explorer, Data Warehouse) are hosted in GCP. We are also working on publishing Data Warehouse and Block Explorer in GCP Marketplace. As we continue to develop the Guru Network, we remain committed to utilizing the best tools and resources available, whether in the cloud or on-premises, to deliver unparalleled performance and reliability. We also explore AWS and Azure partnership options now.
Looking forward, Guru is poised to redefine decentralized AI orchestration. Our platform will empower individual agents to operate independently, supporting a wide range of products and features across our ecosystem. By leveraging our open-sourced infrastructure, we aim to create a thriving community of developers and users who can contribute to and benefit from advancements in AI and blockchain technology.
The Guru Network is not just a platform; it is a vision for the future of decentralized AI orchestration, where infrastructure, individual agents, and orchestration nodes work together to create powerful, efficient, and scalable solutions for real-world problems.
We are thrilled to announce the new infrastructure partnerships we are currently working on and will be unveiling in the near future.
The underlying infrastructure layer is crucial for the future success of any AI and blockchain integration, and Guru Network is well-positioned to lead this transformation.By transitioning from cloud-based infrastructure to an on-premises, censorship-resistant, and decentralized network, we have laid the foundation for a robust Infrastructure-as-a-Service (IaaS) model. This model powers individual agents and orchestration nodes across the Guru Network, supporting the utility of $GURU and enabling a dynamic and thriving ecosystem. Our expertise and innovative infrastructure solutions powers up Guru Network as a provider of Multi-Chain AI/Web3 orchestration.
Video Version:
Our infrastructure journey began with DexGuru, initially utilizing AWS for its scalability but incurring significant costs of up to $170k per month. Seeking greater cost-effectiveness and control, we transitioned to Hetzner Cloud/Dedicated until restrictions as a crypto company forced us to build our own cloud completely at some point. This move marked a strategic shift towards unparalleled cost-effectiveness and performance. More on it: https://mirror.xyz/evahteev.eth/POoVA0GiX9XcaQXGwQekS7o57pWwKARYzEW4uofwX2E
Our meticulously planned on-premises infrastructure includes high-performance computing (HPC), advanced storage solutions, and sophisticated virtualization techniques. Key components include:
Kubernetes (K8s): For managing containerized applications.
Elasticsearch: For powerful search and analytics capabilities.
Clickhouse: An efficient column-oriented DBMS.
Redis: Used for caching and as a message broker.
Blockchain Nodes: The backbone of our data pipeline.
This setup provides predictable expenses, complete control over resources, enhanced performance, and improved scalability, particularly for data-intensive applications like OLAP databases.
Guru infrastructure: https://github.com/dex-guru/guru-infrastructure (We started from open-sourcing virtualization first, more coming)
The economy is quite simple here, and with AI Models/GPU requirements it becomes even better, for example to run proposed chatbots:
https://x.com/xgurunetwork/status/1805696565786362135
We widely used Dell servers already, and let’s go through an example here with GPU models running.
During normal hours compute happens there(paid off within 3 months compared to cloud costs for the same compute). But we can redistribute and auto provision new ones on GCP in case of increased traffic.
Same time it has redundancy and scaling of the Cloud in case of Maintenance/Issues or Increased traffic. That’s the core of the Guru Hybrid infrastructure solution. That also allows for future decentralization where node runners would provide distributed compute using the Guru Infrastructure framework to run it in Clouds or On-Premises and earn $GURU. As the first step we, ourselves operating now in multiple environments and releasing the automation scripts and nodes so anyone would be able to do it in future.
And we are not the ones there. I had discussions with many RPC and Data providers operating with blockchain data over the years. Most of them having own on-premises compute and storage centers, which became regular practice within the space. Additional motivation there was crypto winter when everyone were cutting costs and working on cashflow.
Some Web2 projects are also following the same pattern, like the most loud one:
What future awaits us? When cloud providers:
And we answer:
https://mirror.xyz/evahteev.eth/POoVA0GiX9XcaQXGwQekS7o57pWwKARYzEW4uofwX2E
Now with Guru Network ideas and flow we are positioned for more on that front, decentralized and self incentivized
At the heart of Guru Network's infrastructure are Individual Agents—decentralized, autonomous computational entities that execute specific AI-driven tasks. These agents operate as part of our blockchain business process automation (BBPA) engines. By running open-sourced AI models, they enable a rich ecosystem where products and features can seamlessly share context and threads, enhancing the capabilities of our chatbots, image generation tools, and other AI applications. More on Architecture in light paper: https://gurunetwork.ai/assets/img/litepapper_dexguru_network.pdf
The Flow Orchestrator is a critical component, managing the deployment and operation of BBPA engines and AI processes. It provides a low-code development environment that simplifies the creation and management of complex workflows. This orchestration enhances the efficiency of blockchain applications and ensures that AI models can operate effectively across multiple chains and environments.
Our decentralized infrastructure supports a dynamic IaaS model, empowering developers and businesses to leverage our tools to build innovative applications. By publishing our infrastructure as open-source, we foster a collaborative environment where participants can contribute to and benefit from the advancements in AI and blockchain technology.
The $GURU token underpins the economic operations of the Guru Network, serving as a fundamental element of our IaaS model. It incentivizes node runners, contributors, and users, facilitating a thriving ecosystem. Each transaction and process within the network is compensated in $GURU, ensuring a stable and predictable economic environment.
But with all that motivational speech, we should say CLOUD PROVIDERS are crucial for interoperability, CDN level solutions, as well as try fast - deploy wisely. In our Hybrid infrastructure solution they just no occupying every level. Our partnership with Google Cloud has been useful. The GCP program for Web3 startups allowed us to experiment with different combinations of GPU/LLM Models servers and optimize our data distribution layer for performance. As we call it “Hybrid solution”: the compute layer(costly and heavy one) stays within our On-Premises DC, while APIs and FE for B2B clients (Block Explorer, Data Warehouse) are hosted in GCP. We are also working on publishing Data Warehouse and Block Explorer in GCP Marketplace. As we continue to develop the Guru Network, we remain committed to utilizing the best tools and resources available, whether in the cloud or on-premises, to deliver unparalleled performance and reliability. We also explore AWS and Azure partnership options now.
Looking forward, Guru is poised to redefine decentralized AI orchestration. Our platform will empower individual agents to operate independently, supporting a wide range of products and features across our ecosystem. By leveraging our open-sourced infrastructure, we aim to create a thriving community of developers and users who can contribute to and benefit from advancements in AI and blockchain technology.
The Guru Network is not just a platform; it is a vision for the future of decentralized AI orchestration, where infrastructure, individual agents, and orchestration nodes work together to create powerful, efficient, and scalable solutions for real-world problems.
We are thrilled to announce the new infrastructure partnerships we are currently working on and will be unveiling in the near future.
No activity yet