Full catalog
Search 中文
More
AI VISTA / SITE NAVIGATION

Courses and topics open directly from the main navigation.

Open the full content catalog

Courses

4
  1. 01Build your first testable AI workflow5 lessons
  2. 02Make AI answers reliable enough to use14 lessons
  3. 03Design answers that can show their evidence11 lessons
  4. 04Automate with permission, recovery, and cost controls37 lessons
View all courses →

Topics

9
  1. 01Task framing4 guides
  2. 02Context and prompts6 guides
  3. 03Evaluation7 guides
  4. 04Grounded answers11 guides
  5. 05Agent architecture12 guides
  6. 06Agent safety5 guides
  7. 07Local model deployment7 guides
  8. 08Model operations7 guides
  9. 09Field cases4 guides
View all topics →

AI VISTA / COMPLETE KNOWLEDGE CATALOG

Every course, topic, and guidein one place.

Choose a course for a connected path, a topic for a specific decision, or a guide when you already know the task. This page lists every published entry.

4courses 9topics 63guides

01 / COURSE DIRECTORY

4 complete learning paths

Each course has a syllabus, working artifacts, and a clear acceptance check. Enter where your task begins.

  1. 01
    60–90 minutes · 5 lessons

    Build your first testable AI workflow

    Move from an attractive idea to a bounded task, a workflow-or-agent decision, and a ten-case pilot a colleague can review before further investment.

  2. 02
    2–3 hours · 14 lessons

    Make AI answers reliable enough to use

    Preserve the right context, define content and format gates, test untrusted and difficult inputs, then diagnose failures in evidence, instructions, parsing, or handoff.

  3. 03
    2–3 hours · 11 lessons

    Design answers that can show their evidence

    Choose an answer method, approve source versions, then test chunking, retrieval, access, index updates, citations, freshness, and no-answer behavior.

  4. 04
    Core 3–4 hours; branches vary · 37 lessons

    Automate with permission, recovery, and cost controls

    Build a bounded tool-using workflow with control ownership, data and action authority, stop rules, human review, recovery, service fallback, cost, and drift checks. Test delegation, MCP, memory, models, and deployment only when your task needs them.

02 / TOPIC DIRECTORY

9 decisions, one closest to your task

Topics gather related guides around one decision. Choose the question that matters most right now.

  1. 01
    4 guides

    Task framing

    What exactly should AI do—and should it do it at all?

  2. 02
    6 guides

    Context and prompts

    What must the model see and return for the next step to work?

  3. 03
    7 guides

    Evaluation

    What evidence would make this workflow safe to expand?

  4. 04
    11 guides

    Grounded answers

    When does external evidence improve the answer enough to justify a retrieval system?

  5. 05
    12 guides

    Agent architecture

    Who owns each agent step, and how can a reviewer verify what actually happened?

  6. 06
    5 guides

    Agent safety

    How much authority can this agent safely hold?

  7. 07
    7 guides

    Local model deployment

    Can this model serve our workload on local hardware, and when is that deployment worth it?

  8. 08
    7 guides

    Model operations

    Which model makes this workflow dependable and affordable in operation?

  9. 09
    4 guides

    Field cases

    What actually changes when an AI workflow meets a real review, a source update, or a payment?

03 / GUIDE DIRECTORY

Every original practice guide

Grouped by topic, with each title and summary visible. Every guide appears once and opens directly.

01 / 4 GUIDES

Task framing

View topic ↗
  1. 01
    Should this task use AI at all?Judge a task by variation, verifiability, stakes, data boundaries, and fallback before choosing a model or building a workflow.
  2. 02
    Turn an AI idea into a testable task briefReplace a vague AI feature request with a one-page brief that names the user, input, acceptable output, evidence, boundaries, and fallback.
  3. 03
    Workflow or AI agent? Choose the control loop before buildingDecide when a fixed AI workflow is enough and when an agent needs bounded control, tool feedback, stop rules, and a human handoff.
  4. 04
    Draw the data boundary before AI sees the inputRoute every field through allow, transform, isolate, or exclude so a useful workflow does not quietly become an uncontrolled data transfer.
02 / 6 GUIDES

Context and prompts

View topic ↗
  1. 01
    Map the context before rewriting the promptSeparate instructions, source material, examples, history, tools, and output rules to see what the model actually receives.
  2. 02
    How many tokens should one LLM call reserve for input and output?Count the next complete request, reserve room for tool results and all generated tokens, then check model input, output, and context limits before sending it.
  3. 03
    When should you compact a long AI conversation?Set a compaction checkpoint from measured context use, preserve decisions and unresolved actions, then test whether the next turn still follows the facts.
  4. 04
    Write the acceptance contract before the promptDefine required fields, evidence, tone, uncertainty, and rejection conditions before optimizing prompt wording.
  5. 05
    Choose JSON Schema, Markdown, or plain text for AI outputMatch an AI response format to its downstream consumer, then route refusals, truncation, parser errors, schema errors, and factual errors separately.
  6. 06
    Test prompt injection in documents and tool outputsBuild a regression set for untrusted pages, files, retrieval snippets, and tool results; check intended behavior and permission gates before release.
03 / 7 GUIDES

Evaluation

View topic ↗
  1. 01
    Run a ten-case AI pilot before you scaleBuild a deliberately small test set with ordinary, difficult, ambiguous, and unsafe cases so a promising demo becomes evidence.
  2. 02
    Build an AI evaluation set that stays usefulSample real task slices, write reviewable references, calibrate graders, and version the set so model changes produce trustworthy comparisons.
  3. 03
    How do you calibrate an AI grader against human review?Compare rule checks, an LLM judge, and independent human decisions on the same outputs; find false passes by task slice before trusting automated scores.
  4. 04
    How to set AI classification thresholds with a human review queueRoute calibrated binary probabilities into automatic approval, automatic rejection, and human review using error costs and reviewer capacity, then validate the rule on held-out cases.
  5. 05
    How to test LLM temperature and sampling settingsCompare sampling controls on repeated, frozen cases; measure task errors and useful variation without assuming temperature zero guarantees identical answers.
  6. 06
    Diagnose an AI answer before changing the modelTrace a bad answer to missing input, conflicting instructions, weak evidence, capability limits, or a broken handoff.
  7. 07
    The eval that would have caught this: turn one bad output into a regression testA practical regression workflow that turns one production failure into a labeled case, a pass condition, and a gate before the next prompt or model change.
04 / 11 GUIDES

Grounded answers

View topic ↗
  1. 01
    Prompting, RAG, or fine-tuning? Choose by failure modeDiagnose the failure first, then compare instructions, request-time retrieval, and fine-tuning on the same holdout cases and operating gates.
  2. 02
    Decide whether RAG is worth buildingCompare knowledge volatility, source volume, citation needs, access rules, and maintenance cost before adding retrieval.
  3. 03
    How do you approve sources before adding them to a RAG knowledge base?Decide which exact documents and revisions may ground a RAG answer by checking authority, ownership, use rights, audience, effective dates, and conflicts before indexing.
  4. 04
    How do you validate PDF, table, and image extraction before RAG indexing?Compare parsed content with original pages and cells before chunking. Check OCR, reading order, table relationships, figures, and source locations.
  5. 05
    How should you chunk RAG documents? Keep the full evidence intactCompare fixed windows, overlap, and document structure on clauses and tables, then choose chunk boundaries using complete evidence, citations, and cost.
  6. 06
    Keyword, vector, hybrid, or reranking for RAG? Choose with a matched trialCompare BM25, dense vectors, hybrid fusion, and reranking on exact IDs, paraphrases, and mixed questions while holding evidence, access, context, and latency gates fixed.
  7. 07
    Evaluate RAG retrieval before judging the answerBuild a small retrieval trial that checks gold passages, no-answer questions, document versions, ranking, and access boundaries before generation.
  8. 08
    Document-level access control for enterprise RAG, from search to citationCarry verified identity, tenant scope, and document ACLs into every chunk, then test denial across retrieval, model context, citations, cache, and revocation.
  9. 09
    How do you verify that a RAG document update or withdrawal reached the index?Trace a source revision or withdrawal through ingestion, chunk and vector cleanup, search visibility, caches, and answer probes before calling a RAG knowledge base current.
  10. 10
    When RAG lacks evidence, should it answer, clarify, or hand off?Route a RAG request to a supported answer, a specific clarifying question, an honest unable-to-verify response, or human review when retrieved evidence is incomplete.
  11. 11
    Test citations for support and freshnessAudit whether each important claim is supported by the cited passage, uses the right version, and survives a freshness check.
05 / 12 GUIDES

Agent architecture

View topic ↗
  1. 01
    Design an AI agent tool contract before exposing the functionSpecify an agent tool's input schema, preconditions, authorization, side effects, idempotency, timeout results, and trace before making it callable.
  2. 02
    How to implement an AI agent tool-calling loopBuild a provider-aware round trip that assembles complete calls, validates before execution, pairs every result, and resumes from durable state.
  3. 03
    When should an AI agent continue, retry, or stop?Use observable progress, separate turn and cost budgets, error classes, and duplicate-action checks to govern each step of an agent run.
  4. 04
    Subagent or multi-agent? Choose the delegation topologyCompare one agent, a manager calling specialists, a conversation handoff, and parallel delegation by control ownership, shared state, cost, and failure recovery.
  5. 05
    When should an AI agent save, correct, or delete long-term memory?Design a cross-session memory rule for what to save, how to handle a changed preference, and how to verify deletion across retrieval and retained copies.
  6. 06
    How can a long-running AI agent task resume without repeating work?Design a durable checkpoint, approval and operation ledger, and single-owner resume gate for agent jobs that outlive one process or wait for a person.
  7. 07
    How to debug one AI agent run from its observable traceReconstruct a single agent run with spans, tool call IDs, approval changes, and external state to find the first evidenced divergence without exposing sensitive trace data.
  8. 08
    What should you check before connecting a third-party MCP server?Vet a third-party MCP server's discovered tools, OAuth or local credentials, data destinations, version changes, failure behavior, and revocation before an agent can use it.
  9. 09
    When should a repeated AI task become an agent skill?Decide whether a recurring task has a stable trigger, inputs, steps, acceptance test, and owner before packaging it as a reusable agent skill.
  10. 10
    How do you load and unload an agent plugin safely?Test registrations, event errors, resource cleanup, two hot reloads, disablement, partial initialization, and external effects before shipping an agent extension.
  11. 11
    Your MCP server works locally. What breaks when a team depends on it?A production-readiness review for MCP servers covering per-user credentials, tool allowlists, tenant boundaries, rate limits, audit logs, and failure isolation.
  12. 12
    Log the decision, not the prompt: a trace schema for AI agentsA practical trace schema for agent runs that records identity, tool effects, approvals, state changes, and costs—without storing secrets or full prompts.
06 / 5 GUIDES

Agent safety

View topic ↗
  1. 01
    Give an agent the smallest useful permissionPlace each action on a five-level permission ladder, from read-only suggestions to explicitly approved irreversible work.
  2. 02
    How do you test an AI agent sandbox's file, network, and credential boundaries?Run observable allow-and-deny probes for an agent execution environment's mounts, egress, commands, resource limits, and credentials before it handles untrusted work.
  3. 03
    Put human review where it can change the outcomeChoose review gates from impact, reversibility, and uncertainty, then give reviewers the evidence and actions needed to make a real decision.
  4. 04
    Design recovery before the agent failsWrite the stop signal, saved state, owner, rollback action, and safe retry rule before automation reaches production.
  5. 05
    Claude Code without the chaos: a review-first workflow for coding agentsA team workflow for coding agents that gates the task, the diff, and the runtime before a human merges—without turning review into a rubber stamp.
07 / 7 GUIDES

Local model deployment

View topic ↗
  1. 01
    How much VRAM does a local LLM need? Budget weights and KV cacheEstimate raw model weights and KV cache separately, test context length and concurrency, then measure the missing peak costs on your serving stack.
  2. 02
    Is 4-bit LLM quantization worth it? Test fit, quality, and speedCompare an unquantized BF16 reference with a named 4-bit package on fixed task cases and serving load before trading quality for memory.
  3. 03
    LoRA or QLoRA fine-tuning? Run a small, reviewable pilotAdmit a stable behavior problem, split training data without leakage, probe real training memory, and compare adapters with a frozen baseline on unseen cases.
  4. 04
    Ollama, llama.cpp, vLLM, or SGLang? Choose a local inference engineShortlist a local LLM server by model package, hardware, API behavior, and operating needs, then run the same acceptance and service trial before committing.
  5. 05
    How many concurrent requests can a local LLM serve?Find the highest tested arrival rate that meets first-token, output-speed, queue, error, and memory targets on your actual inference server.
  6. 06
    How many GPUs does a local LLM service need?Turn a tested per-replica arrival rate into a candidate replica count, then verify GPU placement, failure capacity, warm-up time, and rollout peaks.
  7. 07
    Local LLM or cloud API? Choose by data, peak load, and operationsCompare owned hardware, a team-managed rented GPU, and a managed model API using data paths, measured peak capacity, operating ownership, and cost per accepted task.
08 / 7 GUIDES

Model operations

View topic ↗
  1. 01
    Choose a model with a job-specific scorecardWeight quality, latency, tool use, privacy, recovery, and operating constraints using cases from the workflow you will ship.
  2. 02
    How should an AI model service fail over? Plan retries, queues, and human handoffRoute a failed model request by error type, user deadline, data boundary, and an evaluated backup model instead of treating every outage as a retry.
  3. 03
    AI inference latency debugging: trace the wait from request to screenSeparate client, gateway, model, tool, and rendering delays in one request trace before changing a model or adding hardware.
  4. 04
    Map the full cost of an AI workflowCount input, output, retries, tools, waiting, review, and failure recovery instead of comparing a single token price.
  5. 05
    Why prompt caching misses: run a three-request testMeasure cache writes, reads, latency, and accepted-task cost with a stable prefix and a changed-prefix control, using provider-specific rules.
  6. 06
    Monitor an AI workflow before drift becomes an incidentWatch task mix, inputs, dependencies, outcomes, and human overrides against a named baseline so every alert leads to an inspectable response.
  7. 07
    Your AI agent's real cost is cost per accepted taskA cost model that counts tokens, retries, tools, human review, and failure recovery per accepted outcome—not per API call.
09 / 4 GUIDES

Field cases

View topic ↗
  1. 01
    AI support triage: the 10-case pilot that earned a review gateA ten-case support triage pilot that separates queue accuracy from authority and turns a wrong refund promise into a review gate.
  2. 02
    AI content review: the citation gate that stopped a confident wrong answerA content team's citation gate for AI drafts: every material claim needs a current source, version, owner, and a pass, revise, or block decision.
  3. 03
    AI research synthesis: when the source changes after the briefA source-change drill for AI research briefs: re-run retrieval, mark dependent claims stale, and hand off a no-answer note when evidence is missing.
  4. 04
    The approval agent that could move money—until we removed the payment toolAn invoice-approval agent that prepares packets but cannot pay, with approval bound to invoice scope and drills for duplicate and unknown payment outcomes.