Sector file // ai-engineering
AI ENGINEERINGDECODED.
Prompts, RAG, agents & evals
Easy
5 questionsWhat belongs in the system prompt versus the user message?
AI Engineering · 795 views · system-prompt, roles, prompting
What are tokens and why do they matter for cost and limits?
AI Engineering · 540 views · tokens, tokenization, pricing
What are hallucinations and what causes them?
AI Engineering · 303 views · hallucination, grounding, reliability
What is a context window and what happens when you exceed it?
AI Engineering · 172 views · context-window, limits, fundamentals
When should you use batch APIs instead of real-time LLM calls?
AI Engineering · 107 views · batch-api, cost, async
Medium
18 questionsHow does streaming work in LLM APIs and why does it matter for UX?
AI Engineering · 1.1k views · streaming, sse, latency
What is the difference between a base model, an instruction-tuned model and a chat model?
AI Engineering · 1.1k views · base-model, instruction-tuning, rlhf
How do you get reliable structured JSON output from an LLM?
AI Engineering · 1.1k views · structured-output, json-schema, validation
Embeddings in 2025-26: how do you choose model, dimensions, and distance?
System Design · 972 views · embeddings, matryoshka
What is the difference between fine-tuning, LoRA and prompt engineering?
AI Engineering · 841 views · fine-tuning, lora, prompt-engineering
How do you estimate and control LLM costs in production?
AI Engineering · 831 views · cost, tokens, optimization
Tool calling vs function calling vs structured output — what changed in 2025?
System Design · 780 views · tool-calling, function-calling
How do you manage conversations that outgrow the context window?
AI Engineering · 771 views · memory, summarization, context-window
Hard
6 questionsHow should an agent handle tool errors, timeouts and retries?
AI Engineering · 1.2k views · error-handling, retries, agents
How do you sandbox an agent and enforce least-privilege tool access?
AI Engineering · 1.2k views · sandboxing, least-privilege, security
Guardrails for LLM apps: how do you implement input/output safety in 2025?
System Design · 993 views · guardrails, safety
How would you design a production RAG pipeline end-to-end in 2025-26?
System Design · 859 views · rag, architecture, retrieval
Why are BLEU and ROUGE poor metrics for LLM outputs, and what replaced them?
AI Engineering · 766 views · metrics, bleu, rouge
What techniques reduce LLM latency at each stage of a request?
AI Engineering · 241 views · latency, ttft, performance
Seen a AI Engineering question that is not here? Add it and help the next candidate.
Share a question