Sam Aleksov
Software engineer
Actions
After years behind the monitor, collaborations with clients from different fields, infinite requirements and the never ending search for unicorns, I'm here to share with you the good, the bad and the evil of Computer Science.
Your first AI model from scratch
Are you excited to build your very first neural network using only the tools you already know?
The pursuit of running AI models on-premises
Whether it's privacy, regulators, cost, or other concerns, sometimes the only viable option is to have complete control over the AI dependencies in your company. Let's look at the different options for deploying and running AI models locally.
Navigating the sea of generative open source models
Let's explore the beautiful world of generative AI models. Besides just the fun of generating high-quality and useful content, we'll also checkout core architectures, model formats and how to deploy these models in your organisation.
Making dreams come true: leveraging the full power of AI for fun and profits
We all have crazy project ideas and never enough time to actually try them out. Let me share with you my approach to agentic coding, workflows and environments that maximise AI productivity.
Local first, offline and secure apps
Expectations are high when it comes to building reliable and secure user experiences. In this talk we'll get a good overview of all the components involved. From APIs like Crypto, Workers, Assembly and AuthN to the new FileSystem interfaces.
Integrating LLMs
An AI first operations model will make the difference for successful companies, products and people.
We’ll look at different ways of integrating popular LLMs into software solutions and workflows.
Join this talk to learn how to run, augment and fine-tune models to help you make the next best thing.
Integrating AI into your applications and workflows
Every experienced software developer will tell you that the most important and empowering factor for the long term success of software is the tooling that you have around it. We'll discover different ways in which you can become more productive by adopting machine learning methods into your workflow and we'll have a tiny peek at the future of software.
Integrating AI into workflows and systems using MCP
This is a talk about common integration patterns using the Model Context Protocol. It explores the main concepts of the protocol, how to connect it to your productivity software, as well as ways to introduce your own data and tooling around it.
We unlock the true power of large language models by moving away from general predictions heading towards specialized context aware generations and tool intenteractions
Extending GitHub Copilot
Let's explore how we to adapt and enrich Copilot to our doman, provide it with more context and improve output quality.
Developing with AI on the go: From Azure to on-premise containers to the living room
You can do more now, someone already is. You could be building something right now. In this session I'll share my workflow and tools to keep a work life balance while increasing productivity.
Deploying LLMs
As a continuation of our series on machine learning and large language models, we’ll explore different approaches to the distribution and inference of popular LLMs.
Building a company around AI
Lessons learned from trying to run a third place and the natural shift towards AI at the core. We'll explore automation and models that help handle traditional brick and mortar store chores as well as ideas to innovate in the space.
7 years of designing & architecting universal apps with React, React Native & Next.js
Let's have a look at common trends and patterns that have emerged over the years of designing solutions for some of the most interesting and challenging apps & clients. From a huge and variate user base to tight deadlines, exotic requirements and sophisticated graphics or 3D, this stack proves over and over again there is nothing that can't be thrown at it.
9 years of designing & architecting universal apps with React, React Native & Node.js
Let's have a look at the trends and patterns that have emerged out of designing solutions for some of the most interesting and challenging apps & clients. From a huge and variate user base to tight deadlines, exotic requirements and sophisticated graphics or 3D, this stack proves over and over again there is nothing that can't be thrown at it.
Fun with lights
Over this last year I've been researching programmable pixels and different brands of rgb light bulb manufacturers. I'll share with you how we can program these lights with javascript, to create colorful shows, improve the ambient lighting of your home or just have some fun and unleash your creativity.
Introduction to Large Language Models. ChatGPT-4, Copilot and Azure
Let's have a look at how LLMs work, checkout some open-source alternatives and try how GPT-4 based models combine with Azure.
JavaScript Unleashed: WebAPIs, x-platform development and flying drones
I'll share with you how to use JavaScript to control DJI drones for fun and profit.
Let there be light: rfx-Art-Net
Art-Net is a simple, yet very powerful industry standard for controlling devices remotely. It is used anywhere from Hollywood virtual production, to your local concert stage effects, to making homes more welcoming and relaxing with beautiful ambient lights. Unfortunately, there is nothing out there that exposes the vast number of devices to javascript and the browser world. I'll share with you the details of how I built rfx-Art-Net to expose Art-Net and DMX devices over WebSockets, how and why it works.
Lowering the entry barrier to 3D
Programming the GPU is seen as challenging and highly specialized, not to mention the very steep learning curve.
We'll look at a framework that simplifies the consumption and creation of 3D scenes and content.
Making 360 photos part of your next app
360° photos facilitate post-production and open the doors for brand new experiences and creativity. Let's explore how they are created and visualized.
WebGPU: one API to rule them all
Let's checkout the current state of WebGPU, how it performs and the history that led us here.
As a bonus, I'll share with you a clever way to use your newly gained knowledge.
You stole the wrong car
This is the story of how a pet project of mine helped me immediately find and retrieve my stolen vehicle. Ultimately we’ll end up discussing how AirTags work, interfacing and developing with Apple’s Find My network and Nearby Interaction framework and of course potential business ideas. We'll also checkout how you can build your own "airtag"-like device and visualise its location on your own frontend app.
Sam Aleksov
Software engineer
Actions
Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.
Jump to top