<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>byToya — AI Product Owner</title><description>Field feedback, tools and methods on AI agents — by Stéphanie Caumont.</description><link>https://www.bytoya.com/</link><language>en</language><item><title>5 signs you need an AI consultant</title><link>https://www.bytoya.com/blog/en/5-signs-you-need-an-ai-consultant/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/5-signs-you-need-an-ai-consultant/</guid><description>You feel AI could transform your business but don&apos;t know where to start? Here are 5 concrete signals that it&apos;s time to work with an AI consultant.</description><pubDate>Tue, 07 Jul 2026 00:00:00 GMT</pubDate></item><item><title>Why I Don&apos;t Use LangChain in Production</title><link>https://www.bytoya.com/blog/en/why-not-langchain/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/why-not-langchain/</guid><description>LangChain is everywhere in AI tutorials. I tested it, then rewrote everything without it. Here&apos;s why, and what I use instead.</description><pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate></item><item><title>Which LLM to Choose for Your AI Project in 2026: A Practical Guide</title><link>https://www.bytoya.com/blog/en/which-llm-choose-ai-project/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/which-llm-choose-ai-project/</guid><description>Claude, GPT-4, Gemini, Mistral, Llama — how to pick the right model for your project? Costs, performance, use cases. Field guide without lab benchmarks.</description><pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate></item><item><title>RAG with Ollama: Local LLM, Private Data, Zero External API</title><link>https://www.bytoya.com/blog/en/rag-with-ollama-local/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/rag-with-ollama-local/</guid><description>Build a fully local RAG pipeline with Ollama. Embedding, vector store, retrieval — without sending a single byte to an external provider. GDPR-ready by design.</description><pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate></item><item><title>Prompt Engineering in Production: What Changes When It&apos;s Real Code</title><link>https://www.bytoya.com/blog/en/prompt-engineering-production/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/prompt-engineering-production/</guid><description>Writing prompts for a demo is easy. In production with real users, the rules change completely. Here&apos;s what I&apos;ve learned integrating LLMs into real products.</description><pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate></item><item><title>Monitoring an AI Agent in Production: Metrics, Alerts, Drift</title><link>https://www.bytoya.com/blog/en/monitoring-ai-agent-production/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/monitoring-ai-agent-production/</guid><description>A deployed AI agent without monitoring is a time bomb. Which metrics to track, how to detect drift, and what I set up systematically in every production</description><pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate></item><item><title>Integrating Claude API in Production: What I Wish I&apos;d Known Before</title><link>https://www.bytoya.com/blog/en/integrating-claude-api-production/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/integrating-claude-api-production/</guid><description>Hands-on lessons from integrating Claude API into real products. Error handling, costs, streaming, retry logic — the points the docs don&apos;t cover enough.</description><pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate></item><item><title>Freelance AI Developer Rates in 2026: What You Can Expect</title><link>https://www.bytoya.com/blog/en/freelance-ai-developer-rates/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/freelance-ai-developer-rates/</guid><description>How much does a freelance AI developer in France cost? Day rates, mission structure, fixed price vs time &amp; materials. A transparent look at market rates.</description><pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate></item><item><title>Fine-tuning, RAG or Prompt Engineering: When to Use Which?</title><link>https://www.bytoya.com/blog/en/fine-tuning-vs-rag-vs-prompting/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/fine-tuning-vs-rag-vs-prompting/</guid><description>Three approaches to specialize a LLM on your data. How to choose between fine-tuning, RAG and prompt engineering based on use case, budget and maintenance cost.</description><pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate></item><item><title>FastAPI + LLM: Deploying an AI API in Production</title><link>https://www.bytoya.com/blog/en/deploy-ai-api-fastapi/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/deploy-ai-api-fastapi/</guid><description>Build and deploy an AI API with FastAPI — async structure, LLM error handling, rate limiting, streaming. What I actually use in production.</description><pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate></item><item><title>Building a Custom MCP Server: Practical Guide</title><link>https://www.bytoya.com/blog/en/building-custom-mcp-server/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/building-custom-mcp-server/</guid><description>MCP (Model Context Protocol) lets Claude talk directly to your APIs and databases. How to build an MCP server in TypeScript, step by step.</description><pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate></item><item><title>Building an AI Agent for Production: The Steps Nobody Shows You</title><link>https://www.bytoya.com/blog/en/building-ai-agent-production/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/building-ai-agent-production/</guid><description>Practical guide to building an AI agent that runs in production. Architecture, tools, testing — what changes between a prototype and an agent that</description><pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate></item><item><title>AI &amp; GDPR: Handling Sensitive Data in Production</title><link>https://www.bytoya.com/blog/en/ai-gdpr-sensitive-data/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/ai-gdpr-sensitive-data/</guid><description>LLMs and GDPR aren&apos;t incompatible — but they require explicit architectural choices. EU hosting, anonymization, on-premise</description><pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate></item><item><title>AI Automation with Python: What Actually Works in Production</title><link>https://www.bytoya.com/blog/en/ai-automation-python/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/ai-automation-python/</guid><description>Python + LLMs for real-world automation — extraction, classification, generation. What I use, what I avoid, and why complex frameworks slow you down.</description><pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate></item><item><title>ChatGPT vs Claude: which one to choose for your AI projects?</title><link>https://www.bytoya.com/blog/en/chatgpt-vs-claude-for-po/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/chatgpt-vs-claude-for-po/</guid><description>An honest comparison of the two dominant LLMs, seen from the field of an AI Product Owner.</description><pubDate>Sun, 15 Jun 2025 00:00:00 GMT</pubDate></item><item><title>Gemini in AI agents: better (and worse) than competitors</title><link>https://www.bytoya.com/blog/en/gemini-for-ai-agents/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/gemini-for-ai-agents/</guid><description>Google Gemini 1.5 Pro in practice: long context, multimodality and Google Workspace integration — what it really changes.</description><pubDate>Tue, 10 Jun 2025 00:00:00 GMT</pubDate></item><item><title>LLaMA, Mistral, Qwen: why open source LLMs are changing the game</title><link>https://www.bytoya.com/blog/en/llama-mistral-open-source/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/llama-mistral-open-source/</guid><description>Open source models have become seriously competitive. What this changes for your AI agent projects — costs, privacy, deployment.</description><pubDate>Thu, 05 Jun 2025 00:00:00 GMT</pubDate></item><item><title>The AI Product Owner&apos;s tech stack in 2025</title><link>https://www.bytoya.com/blog/en/ai-po-tech-stack/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/ai-po-tech-stack/</guid><description>Which tools, which LLMs, which platforms to work effectively as an AI PO? My setup after 6 months in the field.</description><pubDate>Mon, 02 Jun 2025 00:00:00 GMT</pubDate></item><item><title>Why I left dev to become an AI Product Owner</title><link>https://www.bytoya.com/blog/en/why-i-left-dev/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/why-i-left-dev/</guid><description>My journey from full-stack developer to AI PO: the turning point, the doubts, and why I have no regrets.</description><pubDate>Sun, 01 Jun 2025 00:00:00 GMT</pubDate></item><item><title>RAG explained for a Product Owner: when and why to use it</title><link>https://www.bytoya.com/blog/en/rag-explained-for-po/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/rag-explained-for-po/</guid><description>Retrieval-Augmented Generation demystified. What it really is, when it solves a real problem, and when it&apos;s overengineering.</description><pubDate>Wed, 28 May 2025 00:00:00 GMT</pubDate></item><item><title>What 10 years of code taught me about AI agent specs</title><link>https://www.bytoya.com/blog/en/what-dev-taught-me-about-ai-specs/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/what-dev-taught-me-about-ai-specs/</guid><description>Classic AI spec mistakes seen by a developer. And how to avoid paying for them in weeks of dev.</description><pubDate>Mon, 26 May 2025 00:00:00 GMT</pubDate></item><item><title>AI PO vs classic Product Manager: what really changes</title><link>https://www.bytoya.com/blog/en/ai-po-vs-classic-pm/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/ai-po-vs-classic-pm/</guid><description>The AI Product Owner role isn&apos;t just a PM with a prompt on their resume. Here are the real differences, competency by competency.</description><pubDate>Thu, 22 May 2025 00:00:00 GMT</pubDate></item><item><title>My daily Claude Code setup — what actually works</title><link>https://www.bytoya.com/blog/en/my-claude-code-setup/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/my-claude-code-setup/</guid><description>How I use Claude Code in my daily practice: configuration, workflows, pitfalls to avoid.</description><pubDate>Mon, 19 May 2025 00:00:00 GMT</pubDate></item><item><title>Fine-tuning vs prompt engineering: how to choose?</title><link>https://www.bytoya.com/blog/en/fine-tuning-vs-prompt-engineering/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/fine-tuning-vs-prompt-engineering/</guid><description>When is it better to fine-tune a model rather than optimize your prompt? The answer isn&apos;t what you&apos;d expect.</description><pubDate>Thu, 15 May 2025 00:00:00 GMT</pubDate></item><item><title>How to evaluate an AI agent before going to production</title><link>https://www.bytoya.com/blog/en/evaluating-an-ai-agent/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/evaluating-an-ai-agent/</guid><description>The metrics, tests and methods to know if your agent is truly ready — before your users find out.</description><pubDate>Mon, 12 May 2025 00:00:00 GMT</pubDate></item><item><title>Going freelance as an AI PO: what I wish I&apos;d known</title><link>https://www.bytoya.com/blog/en/first-steps-freelance-ai-po/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/first-steps-freelance-ai-po/</guid><description>The real questions to ask before launching, how to find your first clients, and mistakes to avoid when starting out.</description><pubDate>Thu, 08 May 2025 00:00:00 GMT</pubDate></item><item><title>MCP in practice: what it really changes for your agents</title><link>https://www.bytoya.com/blog/en/mcp-practical-guide/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/mcp-practical-guide/</guid><description>The Model Context Protocol explained without jargon, with concrete integration examples and what it changes in your agent architecture.</description><pubDate>Mon, 05 May 2025 00:00:00 GMT</pubDate></item><item><title>Multi-agent orchestration: patterns and pitfalls to avoid</title><link>https://www.bytoya.com/blog/en/multi-agent-orchestration/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/multi-agent-orchestration/</guid><description>How to architect multiple collaborating AI agents? Patterns that work in production and those that create more problems than they solve.</description><pubDate>Thu, 01 May 2025 00:00:00 GMT</pubDate></item><item><title>Prompt Engineering for Product Owners: What Actually Changes</title><link>https://www.bytoya.com/blog/en/prompt-engineering-for-po/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/prompt-engineering-for-po/</guid><description>You don&apos;t need to be a developer to write good prompts. What every PO should know about how LLMs interpret instructions.</description><pubDate>Mon, 28 Apr 2025 00:00:00 GMT</pubDate></item><item><title>MCP explained simply: what it changes for devs</title><link>https://www.bytoya.com/blog/en/mcp-explained/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/mcp-explained/</guid><description>The Model Context Protocol demystified, with concrete integration examples.</description><pubDate>Mon, 28 Apr 2025 00:00:00 GMT</pubDate></item><item><title>Generative AI by sector: where it really creates value in 2025</title><link>https://www.bytoya.com/blog/en/generative-ai-by-sector/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/generative-ai-by-sector/</guid><description>Finance, healthcare, legal, e-commerce, HR — in which sectors is generative AI keeping its promises and where is it still disappointing.</description><pubDate>Thu, 24 Apr 2025 00:00:00 GMT</pubDate></item><item><title>From dev to AI Product Owner: what I wish I&apos;d known</title><link>https://www.bytoya.com/blog/en/dev-to-ai-po/</link><guid isPermaLink="true">https://www.bytoya.com/blog/en/dev-to-ai-po/</guid><description>The transition, beginner mistakes, and the skills that actually make the difference.</description><pubDate>Thu, 10 Apr 2025 00:00:00 GMT</pubDate></item></channel></rss>