Lochin Shodiev

Choosing the right tools and packages can make a big difference in your projects. Here are some of my favorites that I use regularly: 1. ramda 2. generic-pool 3. nestjs 4. kafkajs 5. mongoose 6. luxon 7. opossum 8. jest 9. playwright 10. ajv 11. cdktf 12. commander 13. ulidx 14. ioredis What about you? What are some of your go-to tools and packages that you simply can't do without? #nodejs #softwareengineering

٧

One of the most interesting talks at the AWS Summit Dubai this week was "FMOPS/LLMOPS: Generative AI from Idea to Production on AWS" by Dr. Sokratis Kartakis. The slide below, in particular, was very useful to me. It described what an engineering team might look like with the advent of LLMs. The diagram shows: - Providers: Research institutions hiring AI specialists to build and train foundational models like ChatGPT, Bedrock, Claude, and others. - Fine-Tuners: Data and ML Engineers at a company who can choose the best models for a given use case and fine-tune them as necessary. - Consumers: These could be the backend engineers who feed user input and context into prompts that are then sent to the fine-tuned model. I found this diagram particularly useful because, as a Backend Engineer, it became clear to me that I need to upskill. There are many tutorials and courses available, but it's often unclear where to begin and how deep one should go. The slide suggests that Backend Engineers should at least be familiar with Prompt Engineering and Retrieval Augmented Generation (RAG). For those interested in going deeper, they could learn about how to fine-tune a foundation model. For this purpose, the course "Generative AI with Large Language Models" by Coursera seems the most promising. I hope this post helps anyone who is in the same situation as I was. I'd like to thank Dr. Sokratis Kartakis for his insightful talk.

٤٢

٤ تعليق

N11, a leading Turkish e-commerce platform, safeguards its millions of customers by migrating to Elastic Security on Google Cloud. The results? Massive savings (75% in licensing costs!), a lightning-fast migration with Elastic's partner Secure Computing, AND seamless integration with Google Cloud logs for enhanced visibility. ⚡ Now, N11 enjoys improved security, cost-efficiency, and a unified view of their entire infrastructure. 🛡️💰☁️ #Elastic #GoogleCloud #Security #eCommerce Want to learn more about N11's Elastic journey? Read the full story here:

١

Here is the video of my presentation: One engineer company with Ruby on Rails at Balkan Ruby 👉 https://lnkd.in/dfNByzcr You can also find the slides at https://lnkd.in/dJDxazEH

١٧

⚡ Let's start WeAreDevelopers World Congress 2024 👉 15k developers together in #Berlin 🇩🇪 🗓️ Come to my talk "What's new in Cloud Dev Environment?" on the Google Cloud booth at 11:45 for a quick overview of IDX and Cloud Workstations 💻☁️ #wearedevelopers #developerExperience

٣٠

٨ تعليق

@llama_index people - so the recommended way to build multi-tenancy is to have different query/chat engine object built for each user? even for 100s or 1000s of users. Or is there a way to recreate chat-engine cheap from something? @jerryjliu0 https://lnkd.in/eSaVd668

١

.NET Developers - .NET Minimal APIs let you create HTTP APIs with minimal dependencies. In this video, Milan Jovanović shows how to build a .NET 8 Minimal API on AWS Lambda and Amazon RDS. #dotnet #AWS #serverless https://lnkd.in/ghMf72Cg

٨٠

١ تعليق واحد

🔍 Unlocking the Power of Elasticsearch for Company Name Searches! 🚀 Elasticsearch is powerful yet complex. We realised that it's just hard to get company name searches right 🧩 After numerous tweaks and some creative brainstorming, we've finally found a reliable solution. Check Peakflo (YC W22) Engineering blog to explore the advanced techniques we used to achieve precise, flexible company name searches. Check it out here: https://lnkd.in/ecmqNmf6 Let us know if we can improve it even further. There is always a way to improve things! #Elasticsearch #SearchEngine

٣١

١ تعليق واحد

.NET Developers - In the Event Sourcing pattern, data is stored as a series of events, providing context for app state. In this video, AWS Community Builder Nick Chapsas explains how to get started with event sourcing in .NET with an Amazon DynamoDB example. #dotnet #AWS https://lnkd.in/gVKQ9QCb

١٩

.NET Developers - In RabbitMQ, you can use Headers Exchange to route messages using message headers. In this blog post and video, AWS Community Builder Rahul Pulikkot Nath explores how header exchanges work and shows them in action with Amazon MQ in .NET code. #dotnet #AWS https://lnkd.in/gc6bTiVS

٢٢

🌟 **Optimizing Performance in Modern Architectures with APM Tools** 🌟 In today's fast-paced tech landscape, ensuring optimal performance and reliability is more crucial than ever, especially when dealing with complex systems like microservices architectures and dynamic platforms like Heroku. At ArchLabs, we emphasize the importance of **Application Performance Monitoring (APM)** tools for several key reasons: 🔍 **Complexity Management**: Microservices break down applications into smaller, manageable pieces. However, this modularity introduces complexity in monitoring and managing interactions between services. APM tools offer a comprehensive view, helping to identify bottlenecks across the entire system. ⚙️ **Fault Isolation**: With numerous services interacting, pinpointing the source of an issue quickly is essential. APM tools reduce the mean time to resolution (MTTR) by isolating underperforming services swiftly. 🌐 **Dynamic Environments**: Platforms like Heroku add another layer of abstraction, making traditional monitoring challenging. APM tools provide the necessary insights to monitor performance metrics in real-time, adapting to resource changes dynamically. At ArchLabs, we leverage tools like **New Relic APM** to give our interns real-world experience in: ✅ **Real-time Monitoring**: Tracking performance metrics and transaction traces across microservices. ✅ **High Traffic Handling**: Ensuring applications can manage high traffic loads while maintaining response times under 200 ms. ✅ **Performance Optimization**: Using insights to optimize code and infrastructure for better performance and cost efficiency. By integrating these practices into our internship program, we prepare our future engineers to tackle real-world challenges head-on. Our interns learn not just to develop applications, but to ensure they run efficiently and reliably under any circumstances. Let's build a future where performance and reliability are at the forefront of development! 🚀 #Tech #Microservices #Heroku #APM #NewRelic #PerformanceMonitoring #SoftwareEngineering #ArchLabs #Internship #EngineeringExcellence

٦

Here are the slides for my GopherCon.EU talk: How to Win Frames and Influence Pointers https://lnkd.in/d7byRQfu #gopherconEU #golang

٧٢

٤ تعليق

Last week's DDD Europe was a blast. It seemed architecture was the theme, as session after session touched on it. Many teachable moments and things to revisit once the videos are published. My session on Friday morning was fun - at least for me 😅 We dived into the pros and cons of various approaches to testing collaboration between objects. "In memory" test doubles vs. mocks and stubs, fixtures vs. builders, different approaches to TDD ("London" or "Chicago") and of course: separating logic and context / side-effects entirely. The point was to get and stay in that mode of pros and cons. Test doubles, builders, fixtures, they allow you to progress quickly, but may cause you to suppress feedback on your design choices. Going ham on SRP might cause isolated, simple-to-test code, but at the cost of more files (and more things to name!). It's all trade-offs! The workshop did not go as I had planned. We discussed and debated merits of approaches much more than I had scheduled. The level of experience and insight the attendees brought was amazing, way beyond my expectation. While I'm sad we did not really get into the second assignment, the discussion was "great too" 😉 There are a couple of thoughts I'm left with after the workshop that I'd still like to share: * All the exchanged approaches have their merits and drawbacks; be open about the drawbacks of your approach and try out other styles every now and then! * What level of unit/integration/e2e test suite you're aiming for depends heavily on the project in question. The requirements of the project and the planned properties of the product should feed into your choices on pyramids / ice cream cones / diamonds * One important aspect on testing that was later mentioned in passing in Aki's session is that a unit should be a "black box", its inner workings obscured from the outside, and only its observable behaviour should be under test * Corollary to that; if you experience complexity trying to observe the behaviour of a unit, then that is feedback on your design. * Processing feedback by making it go away (with test doubles, fixtures / builders or other "pain killers") is prohibiting you from designing different / better systems * Relentlessly turning context ("input side effects") into parameters and side effects into return values will eventually lead you to the pattern: load history, apply command to history, produce an amended history. I.e. event sourcing. Some final things to check out: * Growing Object Oriented Software Guided by Tests is a cool book! * The transformation priority premise is a cool thing to check out for TDD aficionados * The SRP can be difficult to actually apply in a discussion (as it is not super objective), but the Integration Operation Separation Principle might be easier for you (https://lnkd.in/eZdFSGM2) * Before jumping to event sourcing, check out Functional Core, Imperative Shell #ddd #dddeu #software #testing

١٣

🔥 Hot tip The shebang line defines which interpreter should be used to execute a script. For example: ``` #!/bin/fish ``` ``` #! /opt/homebrew/bin/fish ``` This is great, but we cannot send a script defined like that to a friend because their fish executable (or whatever else) can be located under a different path. Use this instead 😉 ``` #!/usr/bin/env fish ``` Learned from fish docs: https://lnkd.in/e-aN-vsM #programming 

٥

.NET Developers - You can build generative AI-infused .NET applications by leveraging the foundation models (FMs) supported by Amazon Bedrock. In this post, learn the basics of integrating Amazon Bedrock in your .NET applications. 🧠 https://lnkd.in/dkSpQArr #dotnet #AWS #AmazonBedrock

٤٠

Apache NetBeans is one of the first IDEs for Java. It has been supported for nearly 30 years. Recently, Apache NetBeans 21 was released. We've decided to check the source code of such a long-lived project and have chosen the most curious errors. Let's follow them up in this article. #ApacheNetBeans #Java #OpenSource

٥

.NET Developers - The first preview of AWS SDK for .NET V4 is now available. V4 provides improved performance and additional library NativeAOT compatibility, as well as changes regarding nullable value types and collection return types that you'll want to be aware of. We invite developers to review V4 Preview 1 and provide feedback. Read all about the V4 SDK in Norm Johanson's blog post. 👀 https://lnkd.in/gqEi5j6q #dotnet #AWS

٦١

Hey there, my fellow Java Developers! I am glad to announce the release of Ollama4j, a library that unites Ollama (an LLM manager and runner) and your Java applications! 🌐🚀 👉 GitHub Repository: Ollama4j on GitHub (https://lnkd.in/gShPqwkR) 🌟 Key Features: - Easy integration with Ollama, enabling the execution of large language models locally. - Clean and simple APIs, focused on seamless interaction with Ollama. - Empowers Java developers to harness the full capabilities of Ollama. - Provides APIs to perform operations such as listing, pulling, deleting models, and creating custom models. - Provides APIs to ask questions (generate completions) to the LLMs in synchronous and asynchronous modes. - Ability to ask questions along with image files or image URLs! 🤩 - Open-source and primed for collaborative contributions from the community! 🦙 What is Ollama? Ollama is an advanced AI tool that allows users to easily set up and run large language models locally (in CPU and GPU modes). With Ollama, users can leverage powerful language models such as Llama 2 and even customize and create their own models. For more details about Ollama, check these out: - https://ollama.ai/ - https://lnkd.in/g8MQqnMT 👨‍💻 Why Ollama4j? As a Java developer passionate about harnessing the latest advancements in AI, I realized the need for a simple and efficient way to integrate Ollama into Java applications. That's why I authored Ollama4j! 🔧 How to Get Started: Visit the GitHub repository: Ollama4j on GitHub. Follow the easy setup instructions in the README to integrate Ollama into your Java projects. Start unlocking the potential of large language models in your applications! 🙏 Contributions Welcome: I invite the Java developer community to explore, use, and contribute to Ollama4j. Your feedback, suggestions, and contributions will help this library get better. I am excited about the possibilities that Ollama4j opens up for Java developers. Whether you're working on natural language processing, chatbots, or any application that can benefit from advanced language models, Ollama4j is here to elevate your projects. I look forward to seeing the incredible applications/projects you'll build with Ollama4j! 🌟 You can get full API details here: https://lnkd.in/gzmWTX7h #Java #AI #LLM #GenAI #GenerativeAI #Ollama #Ollama4j #OpenSource #Developers

٥١