Meet Shai Ben Shalom, Senior Software Engineer at Torque (ex-Microsoft)

Theme: AI-Assisted Enterprise Coding (and hackathons...)

Highlights from the conversation:

• Your job just changed completely: Developers are becoming "mini-architects" instead of code writers. AI writes the code now, but you need to focus on security, scalability, and system design to make it production-ready.
• Put everything in Git for smarter AI: Keep all docs and code in one place so AI agents can understand your full project. Shai uses specialized agents for code reviews and testing that know the entire codebase context.
• Platform engineering makes you unstoppable: Learning the full stack from Kubernetes to databases eliminates dependencies. You can deploy and fix anything without waiting for other teams to help you.
• The gap isn't technical anymore, it's imagination: APIs and tools are ready - the question is how creative you can get with what problems to solve. From natural language database queries to AI troubleshooting your infrastructure.

Enjoy!

Meet Moshik Shir - Principal Program Manager for Startups and Software Development at Microsoft, working across EMEA with 25+ years in tech, covering everything from VoIP pioneers to Azure resilience engineering, in a conversation about AI-Assisted Enterprise Coding.

Highlights from the conversation:

Context is everything - developers who are good at giving AI detailed, specific information become huge fans of AI coding. Those who give vague prompts end up thinking AI sucks. Context management is a skill you can learn and it makes all the difference.

AI agents talking to other AI agents - we're moving from humans using AI to AI agents working together. One agent books your flight while another handles your hotel. This will completely change how companies handle support and knowledge management.

Microsoft's winning combo - GitHub Copilot for coding + Microsoft Research (like Perplexity) for starting new projects + Teams Copilot for meeting notes. Using the research tool with Copilot together is especially powerful when kicking off new work.

Learn two things, not just one - being super specialized in one area is risky now. Build skills in two domains (like tech + business, or frontend + AI) so you can adapt when AI changes your job.

Copilot is everywhere at Microsoft - it works in Excel, Teams, Azure, VS Code - basically everything. People use it for finding old files, prepping for customer meetings, and getting meeting summaries without needing an assistant.

Enjoy!

Meet Kieran Llarena - a hacker, startup/community founder, Capital One intern, and co-founder @ Filipino Americans!

In this episode, we talked about AI-assisted enterprise coding. Context management is one of the most critical keys to success when working with AI (LLMs) - and in this episode, Kieran shared his experiences and best practices.

We dove into his favorite tools and why Kieran loves them. Here are some of the tech we talked about:
https://www.warp.dev/
https://www.trynia.ai/
https://zed.dev/
https://www.diabrowser.com/

We also discussed why hackathons are unparalleled when it comes to exploring new game-changing technologies!

Enjoy!

Meet Balaj Saleem - Software Engineer at Amazon working on AI adoption at IMDb, with experience from both startups and big tech, in a conversation about AI-Assisted Enterprise Coding.

Highlights from the conversation:

The 80/20 rule for AI coding - spend 80% of time planning and providing context to the LLM, only 20% implementing. The cost of fixing bad code later far outweighs upfront planning.

Repository-level markdown files are game-changers - create template files with business context, system assumptions, and coding guidelines that LLMs can reference for every prompt in that codebase.

Question-answer sessions before coding - have 15-20 critical questions with the AI to understand the problem fully before writing any code. Often you're not 100% clear on what you want to build.

LLMs excel at writing tests - they're incredibly good at test generation for existing code, making this a low-risk, high-value starting point for AI adoption.

Enjoy!

Meet Rian Corcino - 7-time hackathon winner and AI Systems Support Specialist at Ayzenberg in a conversation about AI-Assisted Enterprise Coding.

Rian shares his experience and insights - some highlights from the conversation:

• 🏆 90% of his professional development came from hackathons - not from school, but from 20 hackathons in one year traveling across the US
• 🎯 His "responsible vibe coding" approach - generates lots of code with Cursor but reads everything and asks for comments to understand how it works
• 🛠️ His AI workflow stack - Claude for architecture planning, Cursor for coding, Warp for command line tasks, and ChatGPT for... therapy sessions
• ⚡ Why AI + hackathons = perfect match - you can now actually build and ship real B2B SaaS products overnight instead of just prototypes
• 📈 His winning strategy - starts with ideation and customer focus, not technology. Adapts presentations based on whether judges are technical or business-focused

Enjoy!

Meet JD Fiscus - Director of Research & Development, creator of the n8n MCP community node with 1 million downloads and host of Nerding I/O Youtube, in a conversation about AI-Assisted Enterprise Coding.

JD shares his experience and insights - some highlights from the conversation:

• 🛠️ Why he always returns to Cursor - loves their auto model feature that picks the right model for each task, keeping costs down while maintaining quality
• 🔒 His golden rule: human in the loop always - never ships code where MCP writes directly to database or GitHub without his final approval
• ⚡ Context engineering is everything - limits tools to 40 max, clears context frequently, and tasks out specific problems instead of zero-shot prompting
• 🌐 MCP servers are undervalued - most people just wrap APIs, but resources, prompts, and bidirectional sampling are the real game-changers
• 📊 Speed metrics are misleading - can't measure velocity the same way because you're working completely differently - focus on creativity and ability to pivot instead

Enjoy!

Meet Ventsislav Petrov, Frontend Developer at Knowify.

Topic: AI-Assisted Enterprise Coding

Ventsislav shares his experience and insights - some highlights from the conversation:

• 🤯 He writes 100% AI code now - only does analysis, architecture and orchestration while AI handles all the typing
• ⚡ Got 4000x performance improvement in 12 iterations - turned an 8-hour data analysis into 4 minutes with the same dataset
• 💸 Burns 50 million tokens in a single day - doesn't care about token limits because the ROI on complex problems is massive
• 📚 Spends more time reading documentation than code - because AI writes so much code so fast that humans can't keep up reviewing it all
• 🎭 Hid his AI usage in the beginning - just like "many developers" who weren't admitting they were using AI tools at work

Enjoy!

Meet Viktor Farcic, Developer Advocate, Upbound.

Topic: AI-Assisted Enterprise Coding. Viktor shares his experiences - that I first found surprising... but when I thought about it... absolutely obvious:

💸 He spends $1000/month on AI tools but saves $10,000 - gets the equivalent of 3 extra developers working for him

🎯 Why he ditched "cheap" AI subscriptions - they limit context to save tokens, giving him worse code quality

🚀 His company has zero spending limits - and it actually costs them less than hiring expense approvers

⚡ How AI transformed his daily work - coding became a "chore" so he can focus on design and architecture

🔥 Why he's more excited about coding than ever: "I cannot explain how much more fun it is now"