Od LLM fundamentov po produkčné multi-agent systémy
Ako fungujú transformery, attention mechanizmus a tokenizácia
2026-03-04 · 5-6hChain-of-thought, few-shot, system prompts a meta-prompting
2026-03-05 · 4-5hAko evaluovať modely, MMLU, HumanEval, cost/quality tradeoffs
2026-03-06 · 4-5hREST vs streaming, token management, rate limiting, error handling
2026-03-07 · 5-6hEmbeddings, vector databases, chunking strategies, retrieval methods
2026-03-08 · 5-6hHybrid search, reranking, query decomposition, RAGAS evaluation
2026-03-09 · 5-6hIntegrácia všetkého z týždňa 1 do produkčného RAG API
2026-03-10 · 6-8hReAct pattern, tool calling, function calling — základy agentov
2026-03-11 · 6-7hChains, agents, graph-based workflows so state managementom
2026-03-12 · 7-8hAgent komunikácia, delegácia, orchestration patterns
2026-03-13 · 7-8hShort-term, long-term, episodic memory, vector stores
2026-03-14 · 6-7hSchema design, error recovery, validation, dynamic tool discovery
2026-03-15 · 6-7hUnit testy pre AI, eval frameworks, red-teaming
2026-03-16 · 6-7hIntegrácia agentov, memory, tools a evaluácie do jedného systému
2026-03-17 · 8-10hPrompt injection, jailbreaking, input/output filtering
2026-03-18 · 6-7hContainer izolácia, resource limits, network policies, secure execution
2026-03-19 · 5-6hHuman-in-the-loop, approval flows, escalation patterns
2026-03-20 · 5-6hLogging, tracing, monitoring LLM calls, cost tracking
2026-03-21 · 5-6hLoRA, QLoRA, kedy fine-tunuť vs prompt engineer
2026-03-22 · 6-7hCircuit breakers, fallbacks, caching, retry strategies
2026-03-23 · 5-6hIntegrácia security, sandboxing, observability a design patterns
2026-03-24 · 8-10hCompliance, bias detection, fairness metrics, EU AI Act
2026-03-25 · 5-6hNahrádzanie workflow agentmi, cost analysis, ROI
2026-03-26 · 4-5hScaling AI, load balancing, A/B testing models, canary rollouts
2026-03-27 · 6-7hVlastné eval suites, measuring quality, continuous evaluation
2026-03-28 · 6-7hDAGs, state machines, error propagation v multi-agent systémoch
2026-03-29 · 6-7hAlignment, safety, transparency, explainability
2026-03-30 · 4-5hBuilding AI portfolio, open source, personal brand
2026-03-31 · 4-5h