Alex
DevStories.EU
DevStories. EU is your audio companion for the new developer landscape. We publish three types of content: in-depth audio courses on frameworks and tools like FastAPI, Docker, LangChain, and many more (10-20 episodes each, built from official docs), quick Pulse episodes covering recent releases and emerging tools, and annual Special episodes with year-in-review recaps and industry predictions. Designed for learning away from a screen — on your commute, during a run, or while doing the dishes. Created and curated by Alex,...
Autor
Alex
Categoría
Web del podcast
Último episodio
8 de abr. de 2026
¿Dónde escuchar?
Podcasts en la app Replaio Radio Muy prontoLos podcasts llegarán muy pronto a la app. Instálala ahora y sé el primero en descubrir una forma totalmente nueva de vivir los podcasts
Episodios
DeepEval: Unit Test Your LLMs Like Standard Code 08.04.2026 2:16
Stop evaluating your newly deployed language models by manually reading spreadsheets and start writing automated unit tests. This full audio course explores how to build a strict testing pipeline with DeepEval, covering RAG hallucination metrics, automated red teaming, and Pytest integration. Listen to the complete, free audio course at devstories.eu .
Pydantic is a Coercion Engine, Not a Validator 08.04.2026 2:20
Most developers treat Pydantic as a strict gatekeeper, but its internal Rust engine is actually built for lightning-fast data coercion. This full audio course explores how to master strict mode, discriminated unions, safe serialization, and custom core schemas to handle complex API payloads. Listen to the complete twenty-episode course for free at devstories.eu .
OpenAPI & Swagger: Let Your Docs Write Your Code 08.04.2026 2:32
Stop debugging live network traffic just to figure out what your server actually expects. This audio course teaches you how to invert your process using design-first API development with OpenAPI and Swagger, covering machine-readable contracts, JSON Schema validation, and automated SDK generation. Listen to the full course for free at DevStories. EU .
Microsoft Copilot Studio: Stop Hardcoding Bot Triggers 08.04.2026 2:22
Stop wasting hours mapping user phrases to hardcoded triggers and let generative orchestration dynamically chain your agent's actions instead. This 15-episode audio course dives deep into enterprise AI architecture, covering Entra ID authentication, the Computer Use tool for legacy UI navigation, and the Model Context Protocol. Listen to the complete audio course for free at devstories.eu .
OpenAI Agents SDK: Intercept Prompts Before the LLM Wakes Up 08.04.2026 2:27
Your highest-reasoning model is also your most expensive, making strict guardrails essential for intercepting bad prompts before the LLM even wakes up. This full audio course explores how to handle production challenges with the OpenAI Agents SDK, covering execution boundaries, serializing state for human-in-the-loop approvals, and tool discovery via MCP. Listen to the complete series for free at...
OpenClaw Gateway: Stop Treating Your LLM Like a Blocked Thread 08.04.2026 2:23
Stop treating your local LLM like a single blocked thread that freezes during heavy computation. This audio course explores how to build an asynchronous, self-hosted AI architecture using OpenClaw Gateway, covering ephemeral sub-agents, secure session isolation, and smart context compaction. Listen to the complete course for free at devstories.eu .
PyTorch: How One Underscore Destroys Backprop 08.04.2026 2:27
Learn why a single in-place operation can silently destroy your computation history, or why adding a softmax before Cross Entropy Loss limits your model's ability to learn. This audio course dives deep into PyTorch mechanics, covering data flow, the autograd engine, and PyTorch 2.0 kernel fusion. Listen to the full course for free at devstories.eu .
Langflow: Visual Prototypes That Ship as Production APIs 08.04.2026 2:23
Stop wasting days translating visual AI concepts into server code when your node graph can execute as your exact backend API logic. This audio course covers architectural patterns for real-world Langflow deployments, including decoupled RAG pipelines, concurrent session management, and headless Docker containerization. Listen to the full course for free at DevStories. EU .
NVIDIA NeMo Guardrails: Block Gibberish Prompt Attacks 08.04.2026 2:30
Prompt injection attacks often look like pure gibberish designed to bypass keyword filters and overload a model's attention mechanism. This audio course explores how to secure generative AI applications using NVIDIA NeMo Guardrails, covering perplexity heuristics, Colang 2.0, and execution rails. Listen to the complete audio course for free at devstories.eu .
Prompt Flow: Stop Print-Debugging Your LLMs 08.04.2026 2:14
When an LLM hallucinates, scattering print statements throughout your code is an impossible way to debug. This audio course teaches you how to bring standard engineering discipline to AI projects using Prompt Flow for observability, bulk testing, and automated evaluation pipelines. Listen to the full course for free at devstories.eu .
LlamaIndex: Build LLM Agents That Don’t Delete Prod 08.04.2026 2:25
If you let an autonomous agent take destructive actions without human permission, your pipeline is a massive liability. This fourteen-part audio course teaches you how to build robust LlamaIndex applications by mastering custom multi-agent planners, human-in-the-loop orchestration, and objective RAG evaluation metrics. Listen to the complete course for free at devstories.eu .
Deep Agents: Turn Fragile Chat Loops Into Durable Assistants 08.04.2026 2:26
Building a custom AI coder usually ends with the model hitting context limits and overwriting critical files with garbage. This audio course explores how to use Deep Agents to build durable, long-running assistants with sandboxed execution and progressive skill disclosure. Listen to the full course for free at DevStories. EU .
BMad Method: Why AI Coding Assistants Fail on Day Three 08.04.2026 2:38
Most AI coding assistants fail on day three because they forget why they wrote the code on day one. This audio course explores the BMad Method, teaching you how to fix context blindness with data sharding, configure adversarial code reviewers, and orchestrate multi-agent debates. Listen to the complete free course at https://devstories.eu .
LangGraph: From Brittle Scripts to Resilient State Machines 08.04.2026 2:23
When a long-running AI agent crashes and wipes out transient memory, you lose both your workflow progress and your API budget. This course explains how to fix fragile workflows with LangGraph by implementing durable execution, human-in-the-loop pauses, dynamic fan-out, and execution time travel. Listen to the complete free audio course at DevStories. EU .
LangChain: Why Your AI Agents Fail Silently in Production 08.04.2026 2:14
If your AI agent fails silently in production, it is often because you are letting conversation history bloat the context window or trusting a language model to police its own output. This full audio course breaks down the architectural patterns needed to build predictable software with LangChain v1.0, covering summarization middleware, deterministic guardrails, and distributed multi-agent network...
CircuitPython: Embedded Systems Without the Compile Step 08.04.2026 2:16
Stop fighting with compilers and device drivers just to test a single line of embedded logic. This seven-part audio course explores how CircuitPython eliminates the toolchain, handles memory bottlenecks on constrained devices, and scales bare-metal code to full Linux environments. Listen to the complete audio course for free at DevStories. EU .
MicroPython: Fitting Python Into 256KB of Flash 08.04.2026 2:35
Standard Python habits like dynamic object creation will instantly crash a constrained microcontroller due to rapid memory fragmentation. This audio course explains the engineering realities of MicroPython, covering direct silicon interaction, surviving strict RAM limits, and writing zero-allocation hardware interrupts. Listen to the full course for free at devstories.eu .
PyCharm: Stop Debugging Python with Print Statements 08.04.2026 2:29
If your code is still littered with print statements to figure out why a value is returning None, you are wasting development time. This five-part audio course explores how to tighten your feedback loop in PyCharm through automated environment isolation, native test execution, and advanced debugging for synchronous and async Python. Listen to the full, free course at devstories.eu .
GeoDjango & PostGIS: Why Spatial Apps Crash Under Load 08.04.2026 2:30
Deploying a spatial web app can lead to silent server crashes if concurrent requests trigger threading clashes deep inside the GDAL C-library. This audio course explores how to architect location-aware backends with GeoDjango and PostGIS, covering spherical trigonometry, spatial indexing, and stable WSGI deployments. Listen to the full course for free at devstories.eu .
Rasterio: Stop Fighting GDAL and Just Use Numpy 08.04.2026 2:21
Native GDAL bindings feel like writing C, and loading massive satellite images into memory will easily crash your system. This audio course teaches you how Rasterio solves this by treating geospatial datasets as Numpy arrays, covering windowed processing, Affine transforms, and polygon masking. Listen to the complete audio course for free at devstories.eu .
Shapely: Processing Spatial Data Without Python Overhead 08.04.2026 2:28
Processing millions of coordinate pairs can quickly grind your Python application to a halt if you rely on standard loops. This eight-part audio course explores the mechanics of Shapely, teaching you how to bypass Python overhead with vectorized functions, evaluate spatial relationships with the DE-9IM matrix, and build high-performance STRtree indexes. Listen to the complete free audio course at...
Folium: Interactive Web Maps Without Writing JavaScript 08.04.2026 2:02
Plotting thousands of spatial coordinates directly can easily crash your browser or force you to write custom JavaScript. This audio course teaches you how to use Folium as a translation layer to generate optimized Leaflet maps entirely from your Python backend, covering lazy loading, GeoPandas integration, and DOM performance management. Listen to the complete course for free at devstories.eu .
GeoPandas: Why Your Spatial Queries Bottleneck Your CPU 08.04.2026 2:04
Stop wasting memory by loading massive shapefiles into RAM just to filter out most of the rows. This audio course explores how GeoPandas operates under the hood, covering coordinate reference systems, R-tree spatial indexes, and complex spatial joins. Listen to all episodes for free at DevStories. EU .
LiteLLM: Prevent $10k OpenAI Bills with an AI Gateway 08.04.2026 2:20
Relying on provider dashboards to catch a rogue script racking up a massive OpenAI bill is a structural failure. This audio course teaches you how to use LiteLLM to standardize requests, configure automated model fallbacks, and enforce strict budget limits right at the gateway. Listen to the complete series for free at DevStories. EU .
Zipline: Why Trading Algorithms Fail in Live Markets 08.04.2026 2:17
Most trading algorithms look brilliant on paper but instantly lose money in live markets due to look-ahead bias. This audio course explores how Zipline forces strict event-driven testing to prevent data leaks, covering stable Conda setups, custom data bundles, and live broker execution. Listen to the complete course for free at devstories.eu .
Podcasts similares
Replaio no es editor de podcasts; los nombres de los programas, las portadas y el audio pertenecen a sus autores y se distribuyen a través de canales RSS públicos