About
With over two decades of experience in software development and six years driving…
Services
Articles by Maxim
-
Are you one of the Azure Heroes?
Are you one of the Azure Heroes?
The cloud developer community is growing and the contributions from this group is impressive. We see an amazing number…
1 Comment -
What is the Mobile Era conference?
What is the Mobile Era conference?
Dear developers! Only one week left before we convert lovely Oslo from just the capital of Norway to the capital of…
3 Comments
Activity
32K followers
-
Maxim Salnikov shared thisDevelopers in Norway, AI-native coding event is coming! Before you log off for the well deserved Easter break, make sure you register for GitHub Copilot Dev Day in Oslo on April 22 (08:30 to 12:30): https://luma.com/flzseodl This event is part of the global GitHub Copilot Dev Days series. These are in person, community-led events designed to help developers understand how AI-native development fits naturally into their daily work. In Norway, we will focus on the latest advancements including: - Agentic SDLC - Copilot CLI and SDK - Managing AI native infrastructure, and more All aimed at helping you streamline real development workflows, from planning to coding to deployment. The event team is working hard on the program, but I can already promise valuable insights, great conversations, the best stickers, prizes, and more. 🎁 Huge thanks to Johan Ludvig Brattås from Enora, and Jasvinder Sadana from Experis Norway for leading the community efforts and making this event happen in Norway. 🙌 Hope to see many of you there! Katie Savage, Mario Kedziora, Jon Jahren, Morten Stange Bye, Haakon Hasli, Liselotte Bjerkvik, Christian Tryti, Else Tefre, Jonas Helin, Adil I., Ömür Sert, Muhammad Daniyal (Dani), Pankaj Agrawal, Majid Hajian #GitHubCopilot #AINativeDevInfra #DeveloperCommunity #GitHubCopilotDevDays #AgenticAI #AIEngineering
-
Maxim Salnikov shared thisI started using GitHub Agentic Workflows on day one. And it quickly became part of how I build and maintain developer tools! Like many folks today, I build AI-native dev tools for myself and share them with the developer community (a post about that is coming). One part of this work is 18 (!) agent skills that I created and maintain. Most of these skills cover emerging APIs and libraries. From the very beginning, my goal was clear: automate everything that can be automated to keep these skills up to date. That is their key value for developers. GitHub Agentic Workflows turned out to be the perfect technology for this. It allows mixing rigid operational flows with prompt based agentic tasks. Every Monday, my GitHub Actions run a set of prompts using the *-skill-update.prompt.md pattern. These prompts explain: - where to grab updated specs - how to prioritize changes - what to update and how The result is a set of pull requests waiting for review. Human-in-the-loop. Always! :) 👨💻 But I did not stop at automation. I'm a big fan of meta AI: - I maintain GitHub Agentic Workflows agent skill (and other 17) with… GitHub Agentic Workflows skill - I release Agent Skill Deploy agent skill (and other 17) with… Agent Skill Deploy skill Yes. Agents maintaining agents - this is where things get really interesting! What could be your agentic workflow scenario? There are plenty of examples and ideas here: https://lnkd.in/eNT35Rdz When getting started, install my GitHub Agentic Workflows skill for a better developer experience: ⌨ npx skills add webmaxru/ai-native-dev --skill github-agentic-workflows Francesco Manni, Morten Stange Bye, Liselotte Bjerkvik, Jaime De Mora, Else Tefre, Haakon Hasli, Christian Tryti, Pankaj Agrawal, Majid Hajian, Muhammad Daniyal (Dani), Daniel Meppiel, Joel Norman, Jonathan "Peli" de Halleux, Don Syme, Ömür Sert, Adil I., Sebastien Le Calvez, 🥑 Aaron Powell, Lee Stott #AINativeDevInfra #GitHubCopilot #AgenticAI #DeveloperTools #DevTools #AgenticWorkflows
-
Maxim Salnikov shared thisSpeaking at European AI & Cloud Summit ECS 2026 (Cologne) has become a valued tradition for me. This time, I'll focus on #cybersecurity & #AINativeDev intersection with my session "Critical Evaluation of AI Coding Assistants: Productivity Gains vs. Security Risks". This session looks at what agentic AI means for developer security in practice. We will break down how autonomous and semi autonomous coding agents fail, where trust in automation goes too far, and why traditional secure coding guidance is no longer enough. The focus is on concrete scenarios teams are already facing and on pragmatic guardrails that keep agents useful without giving them unchecked power. The goal is to help security and engineering teams work with agentic AI in a way that scales productivity while keeping control, visibility, and accountability firmly in place. Can’t wait to share, learn and connect with everyone in May! Microsoft speaker line-up at #ECS2026: Lee Stott, Carlotta Castelluccio, Marco Casalaina, Adam Harmetz, Erica Toelle, 🪄 Robin Rosengrün, Josh Ndemenge, Julia Muiruri, Bethany Jepchumba, Pierre Roman, Marlene Mhangami, Kimmo Forss, Vinícius Apolinario, Tomasz Naumowicz, Shruti Ailani, Milan Kaur, Massimo Bonanni, Maanav Dalal, Garry Trinder, Waldek Mastykarz, Paolo Pialorsi, Bernd Vellguth, Vesa Juvonen, Julie Koesmarno #CommunityRocks #AIDevSecurity #cybersecurity Registrations and more info at https://lnkd.in/e6a-fqrS
-
Maxim Salnikov shared thisAI-native development is powerful. But are we thinking enough about its security implications? 🔐 I had a remarkable time presenting AI-native development with GitHub Copilot to the technical team at reMarkable. (If you have not heard about the paper tablet they build, it is definitely worth checking out). Among many topics we discussed, security in AI-native development became one of the main focus areas. One specific question stood out: - How do you manage risks related to third-party AI-native development primitives? Take agent skills as an example. My mental model looks like this: - Start with trusted suppliers There are many agent skill repositories maintained by both large vendors and individuals. I also maintain a curated set of agent skills myself. - But trust alone is not enough Breaches and supply chain attacks are part of today's reality. A thoughtful security audit of newly acquired agent skills is essential. - Forking creates responsibility Most teams customize skills to ground them in their own context. Once you fork, governance, versioning, and provisioning become your responsibility. My recommendation: Use an Agent Package Manager https://lnkd.in/eBnKD42C or similar tooling for enterprise grade management of AI native development primitives, not only skills. This becomes increasingly important as agentic SDLC evolves. Big thanks to the reMarkable team for inviting me to meet their tech teams. I really enjoyed the conversation and was impressed by the level of AI adoption. Looking forward to our next engagements! Martí Colominas, Sondre Dyvik, Eirik Folkestad, John Arild A. Johansen Morten Stange Bye, Liselotte Bjerkvik, Haakon Hasli, Christian Tryti, Daniel Meppiel, Francesco Manni, Jaime De Mora, Adil I., Else Tefre, Ömür Sert, Jonas Helin #AINativeDevInfra #GitHubCopilot #AgenticAI #DeveloperProductivity #AISecurity #AgenticSDLC
-
Maxim Salnikov shared this🪄 AI agents orchestration is a must-have skill for developers - check these 2 tools: We’re moving beyond using a single AI assistant into coordinating multiple AI agents working together inside your codebase. This shift changes how we build software: - From prompting → to orchestrating - From single responses → to collaborative workflows - From tools → to systems There’s a lot of innovation happening in this space right now, especially in developer tooling. Here are two solid examples worth exploring: 👉 Squad https://msft.it/6040QsfrC Squad introduces the idea of an AI team living inside your repo. You describe a task, and Squad: - Breaks it down - Delegates work across agents - Runs tasks in parallel - Keeps shared memory and decision logs All of this happens within your repository context. In short: it turns Copilot into a coordinated development team, not just a coding assistant. 👉 Azure az-prototype https://msft.it/6041Qsfrh An Azure CLI extension that empowers customers to rapidly create functional Azure prototypes using AI-driven agent teams. It supports three AI providers - GitHub Copilot, GitHub Models, and Azure OpenAI - for intelligent code and infrastructure generation. Curious, are you already experimenting with multi-agent setups? Brady Gaster, Joshua Davis, Morten Stange Bye, Liselotte Bjerkvik, Haakon Hasli, Christian Tryti, Else Tefre, Francesco Manni, Daniel Meppiel, Jaime De Mora, Ömür Sert, Lee Stott, Adil I., Joel Norman, Jonas Helin #GitHubCopilot #AIAgents #AINativeDevInfra #AZPrototype #AgentSquad #GenerativeAI #msftadvocate
-
Maxim Salnikov shared thisA gift 🎁 to every developer navigating AI-native development. Today is my 🎂 birthday, and I wanted to give something back to the community. Assess your GitHub repo AI-readiness with one click: 👉 https://isainative.dev or directly isainative.dev/<owner>/<repo> Modern AI dev tools scan across multiple folders to find prompts, instructions, skills, custom agents, and other AI-native building blocks. The problem? This structure keeps evolving, and it’s easy to lose track as things grow. “Is AI-Native” scans all known locations (based on the latest tooling patterns) and gives you a complete report of: - what’s already there - what’s missing - where everything lives You can also save snapshots and track how the AI-nativeness of any public repo evolves over time. Want to check private repositories? I’ve got you covered: - VS Code extension: search for “is ai native” - GitHub CLI extension: "gh extension install webmaxru/gh-is-ai-native" - Standalone CLI: "npx is-ai-native scan" To stay future-proof, the web app also exposes repo scanning via WebMCP both imperatively and declaratively, so browser-based AI agents can plug in seamlessly. Behind the scenes, a GitHub agentic workflow keeps everything up to date by scanning documentation weekly and opening PRs when config changes are needed. Everything is open source. You can explore how I use prompts, custom agents, instructions, and agentic workflows in practice. Happy birthday, “Is AI-Native”… and to me as well 🙂 Your best birthday gift 🎁 to me? Give the tool a try and share this post for your network :) Morten Stange Bye, Liselotte Bjerkvik, Haakon Hasli, Else Tefre, Christian Tryti, Cornelia Bjørke-Hill, Majid Hajian, Pankaj Agrawal, Muhammad Daniyal (Dani), Jonas Helin, Adil I., Ömür Sert, Francesco Manni, Daniel Meppiel, Joel Norman, Jaime De Mora, John Romaios, Lillie Harris, Jon Jahren, Lee Stott, Martin Woodward #IsAINative #AINativeDevInfra #DeveloperTools #WebMCP #AIDevelopment #VSCodeExtension #OpenSource #AIEngineering #DevTools
-
Maxim Salnikov shared this📣 Releasing AI developer tools I worked on! I often present on on-device AI in the browser and AI-native development infrastructure. The time has come to supercharge the first with the latter. To help developers build AI experiences in an AI-native way, I’ve created 👉 A set of the Web AI Agent Skills https://lnkd.in/eahCWMkV - A high-quality, auto-updated (key official sources, manual review) set of agent skills that covers key Web APIs supporting Web AI - Works with your favorite AI dev tool natively - Easy to install via "npx skills" or "apm install" - instructions included As a member of the W3C Machine Learning Community Group, I closely follow the latest specification updates and curate these skills so you can rely on them being aligned with where the platform is heading. Main agent skills included: - WebMCP Skill - Prompt API Skill - Writing Assistance APIs (Summarizer, Writer, Rewriter) Skill - Proofreader API Skill - Language Detector API Skill - Translator API Skill - WebNN Skill While working on the main skills, I also created and actively use a couple of supporting skills for maintainability and scale, and I’m sharing them as well: - GitHub Agentic Workflows - I use "aw" for automatic weekly updates resulting in creating PRs for suggested skill changes - Agent Package Manager - I maintain repo AI infrastructure with "apm" 💡 The goal is simple: Make it easier to build AI-native web applications by providing ready-to-use, production-minded building blocks. If you’re exploring browser-based AI, agents, or modern AI dev infrastructure, this project is for you. ⭐ Star the repo and share it with fellow developers by reposting this announcement. I would really appreciate your feedback and contributions 🙌 Huge thanks to Minko Gechev for the skill-creator skill - I use it for quality control, and to Daniel Meppiel for the fantastic Agent Package Manager - now, a part of Microsoft Open Source. Morten Stange Bye, Haakon Hasli, Christian Tryti, Else Tefre, Francesco Manni, Majid Hajian, Pankaj Agrawal, Muhammad Daniyal (Dani), Jaime De Mora, Martin Woodward, Lee Stott, Christoffer Noring, Kyle Pflug, Sohum Chatterjee, Andrew Nolan, Patrick Brosset, Victor H., Alex Nahas Christian Liebel, Robert Nyman, Thomas Steiner, PhD, Jason Mayes, Dominic Farolino, Khushal Sagar, Mark Foltz, #AgentSkills #WebAI #AIAgents #WebMCP #WebNN #DeveloperTools #AIEngineering #W3C
-
Maxim Salnikov shared thisWill running AI workloads directly in your browser become mainstream? 🤔 At first glance, the idea might sound strange: why should you use your own laptop or phone hardware for AI features inside a web app such as summarization, image recognition, or speech-to-text? The answer is surprisingly simple: - Your data stays on your device - You are not dependent on network connectivity and service outages - And it's free! When AI runs locally in the browser, the web becomes a truly universal AI runtime. To help move the industry in this direction, I’m active in the W3C Web Machine Learning Community Group and often speak about emerging browser APIs such as WebNN and the Prompt API. Recently I had the chance to share this topic with 400+ participants at #BelTech Conference by Kainos. Here are my main takeaways: - On-device AI in browsers is already possible. Modern browsers expose APIs that allow web apps to run AI workloads directly on user devices. - Choosing the right approach matters. A simple decision tree can help: -- Small, granular tasks → use built-in browser AI capabilities -- Core application features → consider WebGPU, WebNN, or frameworks like Transformers.js for full customization, BYOM, and fine-tuning. - User experience must come first. If models need to be downloaded: -- Ask users before downloading large models -- Be transparent about device resources -- Provide cloud fallbacks only after explicit user consent The web platform has always been about reach and universality. Now it might also become one of the most important runtimes for AI. Curious how others see this trend. Will browser-native AI become a default architecture for many apps? 💬 Kyle Pflug, Patrick Brosset, Andrew Nolan, Sohum Chatterjee, Martin Woodward Jason Mayes, Christian Liebel, Robert Nyman #WebNN #WebAI #PromptAPI #BuiltInAI #WebPlatform #EdgeAI #DeveloperTools
-
Maxim Salnikov shared thisLinkedIn is not just a social network. It can be a career amplifier for customer-facing technologists (Solution Architects, Developer Advocates, Solution Engineers, Technical Strategists, and many more roles). Recently I was ranked #8 influencer in Norway overall by Favikon. I’m not affiliated with them and I don’t know the details of their methodology, but the ranking made me reflect on what actually worked for me. Here is my _opinionated_ set of tips & tricks that helped me turn LinkedIn into a professional amplifier: - Grow your network intentionally. My “homework” after almost every email thread, call, or meeting is sending connection requests. It never hurts. Most people are happy to stay connected and it compounds over time. - Share what you learn from real work. Customer engagements generate a lot of valuable insights: real architectures, challenges, lessons learned. Identify what can be shared externally, ask permission (!), turn it into quality content, and publish it. When appropriate, tag the people you collaborated with. - Add visuals. Posts with photos, diagrams, screenshots, or short videos consistently perform better. If you include a photo of yourself presenting or speaking, treat it as a visual hook, not the main content. The value must still come from the insights you share. - Invest time in writing good posts. Avoid simple link drops. Share your perspective, experience, and takeaways. Yes, it takes time. But thoughtful posts build credibility. - A slightly controversial one: Use AI. Wisely. I use AI to correct grammar and improve fluency, not to write the content itself. I even keep a reusable system prompt so the editing process is fast. (Yes, this post was also AI-improved.) - Leverage internal advocacy tools If your company provides advocacy platforms, start by sharing curated content. Over time, try to become a contributor as well. - Be authentic and remember your role. Be yourself, share your voice and perspective. At the same time, remember that you are also an ambassador and a public face of your employer. If you work in a customer-facing technical role, LinkedIn can dramatically expand your reach and impact. What strategies have worked for you? #LinkedIn #TechLeadership #PersonalBranding #ThoughtLeadership
Maxim Salnikov commented on a post
6d
Thanks everyone for your congratulations! Just saying - we have multiple open positions in engineering, marketing, sales, in Norway, Europe, and worldwide. Check here https://careers.microsoft.com/ and DM me if you are interested in getting referral.
Maxim Salnikov replied to a comment
1w
Bjarke Grønberg I kindly ask Squad's author Brady Gaster to provide detailed answer. To my understanding, technicalities behind Squad team are the same as you mentioned (you start it "copilot --agent squad --yolo" and it actively uses subagents) but the DX is different: you have much more visibility and control over what's going on and how the task is split
Maxim Salnikov replied to a comment
2w
Daniel Meppiel APM... works! And works great! Another thing - AI agent did not immediately pick right commands, flags, etc. from its "internal" knowledge and quick reference scan. This is why this agent skill appeared! :) Hopefully, this will improve agentic DX and make agent's apm usage predictable and solid. Your PRs and comments are highly appreciated!
I don't use "dependencies" in "aw" myself so far (need to look at it) but I have a line about it in aw skill: 4. `dependencies` for APM-managed skills and prompts.
-
Maxim Salnikov liked thisI'm planning to go, maybe someone wants to join?
-
Maxim Salnikov liked thisMaxim Salnikov liked this(Sound on 🔊) As promised our unofficial Voxxed Days Zurich Fan Video 🚀! Voxxed Days Zurich speaker lineup rocked hard 🎸: Natalie, Céline, 👩🏻💻 Marit, Lutske, Janina, Nadine, Jasmin, Jonas, Christian, Patrick, Andres, Dimitris, Faris, Wouter, Fabrizio, Roberto, Giovanni, Kevin, Mario, Scott, Cay, Younes, Radek, Soroosh, Josh, Stefano, Loïc, Vojtech, Connor, Maurice, Dan, Łukasz, 🌩️ Warren, Emanuel, Stéphane, Marc, Maxim, Makan, Alessio, Pawel ... and so many more!
-
Maxim Salnikov liked thisMaxim Salnikov liked thisYou can scroll through updates…or you can be where they actually happen. from global leaders like Orange Cyberdefense, GitHub, Microsoft, JPMorganChase, Red Hat and UEFA. These are the teams building the tools, platforms, and systems that developers around the world rely on every single day. And for two days, they’ll be sharing insights & challenges LIVE on our stages. Piotr Wicherski- QA Lead UEFA Mathew Caplan - Director of International Professional Services at Orange Cyberdefense. Teresa Wu- VP of software engineer at JP Morgan Chase, Maxim Salnikov- AI Dev Tools & Platforms Solution Engineer at Microsoft Marc Nuri - Principal Software Engineer at RedHat Joseph Katsioloudes- Senior Developer Advocate at GitHub and the list continues with more incredible names you can check on our website. 📍 Meet us at Romexpo, B1 Pavilion. 📅 June 3rd–4th, 2026. 🎟️ Tickets available now: https://bit.ly/4pGgZa9
Experience & Education
-
Microsoft
View Maxim’s full experience
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
Licenses & Certifications
Volunteer Experience
-
Google Developer Expert
Google Developer Experts
- Present 10 years 7 months
The Google Developer Experts program is a global network of highly experienced technology experts, influencers, and thought leaders who have expertise in Google technologies, are active leaders in the space, natural mentors, and contribute to the wider developer and startup ecosystem.
-
Industry Expert Lecturer
Norwegian University of Science and Technology (NTNU)
- Present 1 year 2 months
Education
Delivered a lecture and a workshop “Exploring AI coding tools and their impact on software security”
Honors & Awards
-
Certificate of Appreciation by Global Skill Development Council (GSDC)
Global Skill Development Council (GSDC)
Certificate of Appreciation for the session "Prompt Engineering - an Art, a Science, or Your Next Job Title?" delivered at the GSDC Global Prompt Engineering Bootcamp 2024
-
Azure Heroes Microsoft European Virtual Open Source Summit badge
Azure Heroes
A verifiable link to the badge: https://enjinx.io/eth/asset/91213575
-
Microsoft Most Valuable Professional
Microsoft MVP Award Program
Award Category: Visual Studio and Development Technologies
Profile URL: https://mvp.microsoft.com/en-us/PublicProfile/5002950
Languages
-
Norwegian
Professional working proficiency
-
English
Full professional proficiency
-
Russian
Native or bilingual proficiency
Organizations
-
Framsia
Co-Organizer
- Presenthttp://meetup.com/framsia
-
Mobile Meetup Oslo
Organizer
- Presenthttp://meetup.com/Mobile-Meetup-Oslo
-
Angular Oslo
Organizer
- Presenthttp://meetup.com/AngularJS-Oslo
-
Sencha Developers Group Norway
Organizer
- Presenthttp://meetup.com/sencha
-
IXDA Oslo
-
- Present
Recommendations received
1 person has recommended Maxim
Join now to viewOther similar profiles
Explore more posts
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top contentOthers named Maxim Salnikov
12 others named Maxim Salnikov are on LinkedIn
See others named Maxim Salnikov