GetAnalyst Podcast

Real systems.
Real decisions.

Deep, practical conversations for systems analysts, business analysts, and solution architects. Episodes are in Russian; English summaries and translated chapters are provided below.

YouTube Apple Podcasts Spotify Amazon Music

Latest releases

Episode 521h 32m

Async Integration with Webhooks and Message Brokers

Design an integration with an AI video service through three architecture options: a monolith without a broker, a monolith with RabbitMQ, and a microservice solution. Includes failure scenarios and development-task decomposition.

Watch on YouTube Episode materials

English chapter timestamps

00:00Intro
00:18The scenario and speaker introduction
03:01Middle+ async integration task: requirements, API, and mockups
08:17How to begin a real integration task
12:28AI video service API documentation
14:06Synchronous vs. asynchronous: polling and webhooks
22:53Message-broker integration: RabbitMQ and Kafka
30:10Reading the API documentation, continued
38:13Designing the async integration architecture
42:07Option 1: monolith without a broker
57:46Option 2: monolith with RabbitMQ
1:18:20Option 3: microservice architecture
1:21:27When a webhook never arrives: retries and safeguards
1:23:57Canceling an asynchronous process
1:26:43Errors, stalled jobs, and difficult scenarios
1:27:59Decomposing the solution into development tasks
1:32:23Takeaways, materials, and recommendations

Episode 5135m

Vibe Coding for Systems Analysts: Building an App with AI

A real case study of an analyst-built AI tool for database design and SQL. See the app, development stack, prompts, token costs, delivery time, and where AI code generation meets engineering.

Watch on YouTube Episode materials

English chapter timestamps

00:00Building an AI database and SQL tool through vibe coding
01:33From law to systems analysis: meet the guest
03:34Technology stack and product idea
05:19Product demo and onboarding
06:25Application registration
06:52Designing an ER diagram from requirements
09:17Analyzing an existing database from DDL or an ER diagram
12:15AI-assisted database audit
12:55Migrating data between database systems
17:14Generating database test data
19:58Admin panel, analytics, and token usage
20:46System and user prompts
22:41Exporting AI-generated reports
24:25Ideas for developing the tool
26:53Who owns AI-generated code?
28:37Do systems analysts need to code?
29:35How vibe coding may affect an analyst’s career
31:28Development time and key challenges
35:19How analysts can start working with AI and vibe coding

Episode 5058m

How to Build an AI Skill in ChatGPT and Claude

Learn how reusable AI Skills differ from prompts and projects, what goes inside a skill, and how to create one for a systems-integration use case.

Watch on YouTube Episode materials

English chapter timestamps

00:00AI Skills for systems and business analysts
02:44Why ordinary prompts are no longer enough
05:23What AI Skills are and why they matter
06:54REST API requirements example
09:44Working before and after Skills
10:39What an AI Skill contains
14:05UML and REST API skill structures
15:32Why analysts need AI Skills
17:57Analyst tasks worth automating
19:09Using Skills in ChatGPT
22:40Creating Skills in ChatGPT
25:55AI Projects as a free alternative
27:00Combining Projects and Skills
28:40Skills in Claude
29:58Skills in Perplexity
30:54Free access vs. paid plans
32:04Creating a Skill from your knowledge
34:05Demo: integration use-case Skill in Claude
45:52Using AI safely at work
48:41Final Skill structure
50:12Moving Skills between tools
51:52Five steps from creation to use
53:28Limitations of Skills
55:03Skill, Project, MCP, and Tool: the differences
56:35Takeaways and recommendations

Episode 4923m

Why Your RabbitMQ Will Fail: A Production Incident

A message queue looks empty while the RabbitMQ node is drowning in data. This post-incident analysis explains how delayed personalized emails turned the broker into an object store.

Watch on YouTube Episode materials

English chapter timestamps