Скачать или смотреть Episode #3. Building AI apps. Security issues, UX that earns trust, and future of software engineers

A wide-ranging discussion with GenAI Architect at Google Cloud and Udemy's best selling instructor. We cover:

– What changes when you move from API demos to production apps (rate limits, multi-cloud, failover strategies)
– Security challenges: prompt injection, MCP as remote code execution, supply chain risks, least-privilege agents
– How to measure reliability (confidence in AI responses, cost-of-error, online and offline evals)
– Why “code is cheap” now and how that’s reshaping engineering roles
– "Trustworthy" UX patterns: transparent agent steps, generative UI, ambient agents
– The job shifts already happening: juniors struggling to get hired, first-line support and QA under threat
– How AI is lowering the barrier to entry in education and what engineers need to stay relevant

🧑‍💼Guest: Eden Marco   / eden-marco  
🎙️Hosts: Anton Zhuravsky   / anton-zhuravsky-40690124   & Dmitry Volokh   / davolokh  
📍137 Podcast on X: https://x.com/137_podcast

This podcast is brough to you by On The Spot Development. We help startups build R&D teams in Poland and have open roles for engineers. Check out the relevant opportunity:
𝐅𝐨𝐮𝐧𝐝𝐢𝐧𝐠 𝐄𝐧𝐠𝐢𝐧𝐞𝐞𝐫 (𝐀𝐈): https://onthespotdev.com/positions/fo...

0:00 Intro & episode roadmap: meet Eden Marco (Google Cloud, Udemy)
2:55 AI apps: API wrapper debate & prototype vs. production
6:12 Challenges of scaling AI apps (rate limits, multi-cloud, failover strategies)
10:14 Choosing providers: big clouds vs niche providers
12:21 Security as core of production: prompt injection, MCP as RCE, supply chain & least privilege
16:07 Gaps in best practices: open-source frameworks (LangChain), dangerous patterns, security debt, in-house app vulnerabilities
21:32 Do platforms like Firebase/Supabase help security? (partial)
24:00 Reliability & evaluations: confidence in AI responses, cost-of-error metric, datasets & online/offline evals
29:05 Agentic workflows: LangGraph, guardrails, deep agents & balancing autonomy vs control
34:00 UX patterns that build trust: transparent agent steps, generative UI beyond “just another chat”
39:34 New interfaces: ambient agents (inbox model) & dynamic generative UIs
46:24 Engineering shifts: code becomes cheap, bridging roles, fewer juniors hired
55:25 Jobs on the brink: first-tier support, QA automation, shrinking engineering teams
59:29 Education changes: barrier-free entry, non-engineers building with AI
1:04:57 Staying relevant as an engineer: MCP, architecture trade-offs, scaling yourself with agents
1:13:46 Philosophical outlook: new abstraction levels, resilience, self-healing systems
1:18:33 Closing & thanks

#AIinProduction #PromptInjection #LangChain #LangGraph #AIUX #GenerativeUI #TechPodcast #EngineeringFuture