---
title: "Visual Explainers — page 4"
description: "Visual introductions to the engineering concepts behind the tools we use."
---

# Visual Explainers — page 4
> Visual introductions to the engineering concepts behind the tools we use.

## Visual Explainers

### [How to give every agent its own versioned workspace](https://yomu.fyi/visual-explainers/how-to-give-every-agent-its-own-versioned-workspace.md)
- Type: How to
- Tags: [AI Agents](https://yomu.fyi/topic/ai-agents.md), [Artifacts](https://yomu.fyi/topic/artifacts.md), [CI/CD](https://yomu.fyi/topic/ci-cd.md), [Git](https://yomu.fyi/topic/git.md), [Sandboxing](https://yomu.fyi/topic/sandboxing.md)

A five-step workflow for turning agent output into versioned artifacts: scope a repo per agent, branch instead of copying, run in an isolated sandbox, gate changes with push-triggered evals, then deploy or roll back.

### [How to Partition AMD GPUs for Shared AI Compute](https://yomu.fyi/visual-explainers/how-to-partition-amd-gpus-for-shared-ai-compute.md)
- Type: How to
- Tags: [AMD](https://yomu.fyi/topic/amd.md), [GPU](https://yomu.fyi/topic/gpu.md), [Isolation](https://yomu.fyi/topic/isolation.md), [OpenShift](https://yomu.fyi/topic/openshift.md), [Performance](https://yomu.fyi/topic/performance.md), [Virtualization](https://yomu.fyi/topic/virtualization.md)

A five-step how-to for deciding when an AMD Instinct GPU should stay dedicated and when SR-IOV virtual functions, VMs, or sidecar isolation can safely share it.

### [Agentic Development: The Four Parts of a Delegation Loop](https://yomu.fyi/visual-explainers/agentic-development-the-four-parts-of-a-delegation-loop.md)
- Type: What is it?
- Tags: [Agentic Development](https://yomu.fyi/topic/agentic-development.md), [AI Agents](https://yomu.fyi/topic/ai-agents.md), [Coding Agents](https://yomu.fyi/topic/coding-agents.md), [Evaluation](https://yomu.fyi/topic/evaluation.md), [Human-in-the-Loop](https://yomu.fyi/topic/human-in-the-loop.md), [Sandboxing](https://yomu.fyi/topic/sandboxing.md)

Agentic development turns AI assistance into delegated work: a harness runs the agents, skills supply domain knowledge, a sandbox bounds execution, and evals verify the result.

### [Consolidating Agent Workloads on Atlas: One Platform, Three Trade-offs](https://yomu.fyi/visual-explainers/consolidating-agent-workloads-on-atlas-one-platform-three-trade-offs.md)
- Type: Trade-offs
- Tags: [AI Agents](https://yomu.fyi/topic/ai-agents.md), [Atlas](https://yomu.fyi/topic/atlas.md), [MongoDB Atlas](https://yomu.fyi/topic/mongodb-atlas.md), [RAG](https://yomu.fyi/topic/rag.md), [Scalability](https://yomu.fyi/topic/scalability.md)

Consolidating agent memory, retrieval, and governance onto MongoDB Atlas removes sync and vendor layers, but ties scaling and control to one platform. This trade-off view shows the payoff, the cost, and the condition that decides it.

### [What Is a Real-Time Test Loop?](https://yomu.fyi/visual-explainers/what-is-a-real-time-test-loop.md)
- Type: What is it?
- Tags: [Aerospace](https://yomu.fyi/topic/aerospace.md), [Chaos Engineering](https://yomu.fyi/topic/chaos-engineering.md), [Latency](https://yomu.fyi/topic/latency.md), [Observability](https://yomu.fyi/topic/observability.md), [Predictive Maintenance](https://yomu.fyi/topic/predictive-maintenance.md), [Real-time](https://yomu.fyi/topic/real-time.md)

A real-time test loop is the closed path from sensor acquisition to control action and back to verification. This explainer maps its four parts and the latency budget that binds them.

### [Activation Steering: One Dial, Two Costs](https://yomu.fyi/visual-explainers/activation-steering-one-dial-two-costs.md)
- Type: What is it?
- Tags: [Activation Steering](https://yomu.fyi/topic/activation-steering.md), [Evaluation](https://yomu.fyi/topic/evaluation.md), [Fine-Tuning](https://yomu.fyi/topic/fine-tuning.md), [LLMs](https://yomu.fyi/topic/llm.md), [Prompt Engineering](https://yomu.fyi/topic/prompt-engineering.md)

Activation steering edits a model's internal activations to add or remove a concept at inference time. This explainer maps its one control knob, its two costs, and when it beats prompting.

### [Brandwatch AI: Speed Layer vs Trust Layer](https://yomu.fyi/visual-explainers/brandwatch-ai-speed-layer-vs-trust-layer.md)
- Type: Comparison
- Tags: [AI Governance](https://yomu.fyi/topic/ai-governance.md), [Analytics](https://yomu.fyi/topic/analytics.md), [Brandwatch](https://yomu.fyi/topic/brandwatch.md), [Data Visualization](https://yomu.fyi/topic/data-visualization.md), [Human-in-the-Loop](https://yomu.fyi/topic/human-in-the-loop.md), [Social Listening](https://yomu.fyi/topic/social-listening.md)

Compare Brandwatch's AI speed features—Ask Iris, Query Writer, and AI summaries—with the governed checks that make social insights trustworthy before teams act.

### [Application Security as a Closed Loop: Discover, Govern, Protect, Learn](https://yomu.fyi/visual-explainers/application-security-as-a-closed-loop-discover-govern-protect-learn.md)
- Type: Architecture
- Tags: [Application Security](https://yomu.fyi/topic/application-security.md), [Least Privilege](https://yomu.fyi/topic/least-privilege.md), [Security](https://yomu.fyi/topic/security.md), [Threat Intelligence](https://yomu.fyi/topic/threat-intelligence.md), [Vulnerabilities](https://yomu.fyi/topic/vulnerabilities.md), [WAF](https://yomu.fyi/topic/waf.md)

Application security is a closed loop: discover and prioritize risk, govern access and agent behavior, protect at runtime, then investigate and feed findings back into the next cycle.

### [AG-UI as the Agent–UI Contract Layer](https://yomu.fyi/visual-explainers/ag-ui-as-the-agent-ui-contract-layer.md)
- Type: Architecture
- Tags: [A2UI](https://yomu.fyi/topic/a2ui.md), [AG-UI](https://yomu.fyi/topic/ag-ui.md), [AI Agents](https://yomu.fyi/topic/ai-agents.md), [Human-in-the-Loop](https://yomu.fyi/topic/human-in-the-loop.md), [Server-Driven UI](https://yomu.fyi/topic/server-driven-ui.md), [Streaming](https://yomu.fyi/topic/streaming.md)

AG-UI is a typed event contract between an agent backend and a frontend, turning streamed agent state into UI. This diagram maps the layers and shows where human review plugs in.

### [How to Prepare Your Site for the Agentic Browser](https://yomu.fyi/visual-explainers/how-to-prepare-your-site-for-the-agentic-browser.md)
- Type: How to
- Tags: [Bot Management](https://yomu.fyi/topic/bot-management.md), [Browser](https://yomu.fyi/topic/browser.md), [Performance](https://yomu.fyi/topic/performance.md), [Turnstile](https://yomu.fyi/topic/turnstile.md), [WebMCP](https://yomu.fyi/topic/webmcp.md)

A five-step readiness checklist for teams whose sites are increasingly visited by AI agents: expose callable functions, verify bot traffic, measure real-user speed, and control crawler load.

### [How to Make an Astro Content Site Agent-Ready](https://yomu.fyi/visual-explainers/how-to-make-an-astro-content-site-agent-ready.md)
- Type: How to
- Tags: [API](https://yomu.fyi/topic/api.md), [Astro](https://yomu.fyi/topic/astro.md), [Caching](https://yomu.fyi/topic/caching.md), [CMS](https://yomu.fyi/topic/cms.md), [MCP](https://yomu.fyi/topic/mcp.md), [OpenAPI](https://yomu.fyi/topic/openapi.md)

A four-step path to an agent-ready Astro site: content in an Astro CMS, API/CLI/MCP surfaces generated from one schema, and an edge layer that absorbs spikes while agents call functions instead of clicking.

### [Tomcat 11: Lenient vs Strict Defaults, Side by Side](https://yomu.fyi/visual-explainers/tomcat-11-lenient-vs-strict-defaults-side-by-side.md)
- Type: Comparison
- Tags: [Apache Tomcat](https://yomu.fyi/topic/apache-tomcat.md), [Containers](https://yomu.fyi/topic/containers.md), [Deployment](https://yomu.fyi/topic/deployment.md), [Jakarta EE](https://yomu.fyi/topic/jakarta-ee.md), [Java](https://yomu.fyi/topic/java.md), [Migrations](https://yomu.fyi/topic/migration.md)

A side-by-side comparison of legacy Tomcat defaults and the strict defaults in Tomcat 11, showing which behaviors change and what to verify before you upgrade a Java web tier.

### [Buildkite's Hybrid CI Model: Hosted Runners vs Self-Hosted Agents](https://yomu.fyi/visual-explainers/buildkite-s-hybrid-ci-model-hosted-runners-vs-self-hosted-agents.md)
- Type: Comparison
- Tags: [Anka](https://yomu.fyi/topic/anka.md), [Buildkite](https://yomu.fyi/topic/buildkite.md), [CI/CD](https://yomu.fyi/topic/ci-cd.md), [Developer Experience](https://yomu.fyi/topic/developer-experience.md), [macOS](https://yomu.fyi/topic/macos.md), [TeamCity](https://yomu.fyi/topic/teamcity.md)

Buildkite splits CI into a hosted control plane and agents you run on your own machines. This comparison weighs that self-hosted agent model against fully hosted runners so you can pick the right fleet.

### [What Is an AI Factory? Power In, Tokens Out](https://yomu.fyi/visual-explainers/what-is-an-ai-factory-power-in-tokens-out.md)
- Type: What is it?
- Tags: [AI Factories](https://yomu.fyi/topic/ai-factories.md), [Data Centers](https://yomu.fyi/topic/data-centers.md), [GPU](https://yomu.fyi/topic/gpu.md), [Inference](https://yomu.fyi/topic/inference.md), [Model Serving](https://yomu.fyi/topic/model-serving.md), [Performance](https://yomu.fyi/topic/performance.md)

An AI factory is a fixed power and cooling budget converted into model output. This explainer maps its four parts and the throughput-per-watt trade-offs that connect them.

### [Anatomy of a Bug Bounty Pipeline: From First Report to Upstream Patch](https://yomu.fyi/visual-explainers/anatomy-of-a-bug-bounty-pipeline-from-first-report-to-upstream-patch.md)
- Type: Architecture
- Tags: [Bug Bounty](https://yomu.fyi/topic/bug-bounty.md), [HackerOne](https://yomu.fyi/topic/hackerone.md), [Incident Response](https://yomu.fyi/topic/incident-response.md), [Security](https://yomu.fyi/topic/security.md), [Vulnerabilities](https://yomu.fyi/topic/vulnerabilities.md)

An architecture view of a bug bounty program: scope, triage, validation, containment, durable fix, and payout, plus the loop that turns each report into stronger defaults.

### [Where BPF Hooks In: Choosing the Attachment Layer](https://yomu.fyi/visual-explainers/where-bpf-hooks-in-choosing-the-attachment-layer.md)
- Type: Architecture
- Tags: [BPF](https://yomu.fyi/topic/bpf.md), [eBPF](https://yomu.fyi/topic/ebpf.md), [Linux](https://yomu.fyi/topic/linux.md), [Networking](https://yomu.fyi/topic/networking.md), [sched\_ext](https://yomu.fyi/topic/sched-ext.md), [XDP](https://yomu.fyi/topic/xdp.md)

BPF is not one hook but a set of attachment points, and each one sees different traffic. This diagram maps XDP, TC, socket-level BPF LSM, and sched\_ext so you can pick the layer that can actually enforce your rule.

### [Bot Management as an Adaptive Control Loop](https://yomu.fyi/visual-explainers/bot-management-as-an-adaptive-control-loop.md)
- Type: Architecture
- Tags: [Bot Management](https://yomu.fyi/topic/bot-management.md), [Cloudflare](https://yomu.fyi/topic/cloudflare.md), [Load Shedding](https://yomu.fyi/topic/load-shedding.md), [Security](https://yomu.fyi/topic/security.md), [Turnstile](https://yomu.fyi/topic/turnstile.md)

Bot management is an architecture, not a rule list: classify traffic, score it with continuously retrained models, enforce with challenges or blocks, then feed the outcomes back into the model.

### [How a Board of Directors Works: The Governance Control Loop](https://yomu.fyi/visual-explainers/how-a-board-of-directors-works-the-governance-control-loop.md)
- Type: How it works
- Tags: [Auditability](https://yomu.fyi/topic/auditability.md), [Board of Directors](https://yomu.fyi/topic/board-of-directors.md), [Compliance](https://yomu.fyi/topic/compliance.md), [Data Governance](https://yomu.fyi/topic/data-governance.md), [Decision-Making](https://yomu.fyi/topic/decision-making.md), [Engineering Management](https://yomu.fyi/topic/engineering-management.md)

Boards of directors turn ownership into a governance control loop: shareholders elect directors, directors set strategy and guardrails, committees verify independently, and management executes inside those limits.

### [Bloom Filters: The Cheap Gate Before the Expensive Question](https://yomu.fyi/visual-explainers/bloom-filters-the-cheap-gate-before-the-expensive-question.md)
- Type: What is it?
- Tags: [Bloom Filters](https://yomu.fyi/topic/bloom-filters.md), [Caching](https://yomu.fyi/topic/caching.md), [Distributed Systems](https://yomu.fyi/topic/distributed-systems.md), [Performance](https://yomu.fyi/topic/performance.md), [Search](https://yomu.fyi/topic/search.md)

A Bloom filter is a small bit array plus a few hash functions that answers "has this been seen?" with no false negatives and a false-positive rate you choose up front.

### [How Blackwell Serving Turns Peak FLOPs Into Cheap Tokens](https://yomu.fyi/visual-explainers/how-blackwell-serving-turns-peak-flops-into-cheap-tokens.md)
- Type: How it works
- Tags: [Blackwell](https://yomu.fyi/topic/blackwell.md), [GPU](https://yomu.fyi/topic/gpu.md), [Inference](https://yomu.fyi/topic/inference.md), [Mixture of Experts](https://yomu.fyi/topic/mixture-of-experts.md), [Quantization](https://yomu.fyi/topic/quantization.md), [Speculative Decoding](https://yomu.fyi/topic/speculative-decoding.md)

Blackwell inference is a pipeline, not a chip upgrade. This explainer walks five stages — fleet validation, batching, KV caching, low-precision kernels, and speculative decoding — that turn peak FLOPs into cheap tokens.

[Newer explainers](https://yomu.fyi/visual-explainers/page/3.md) · [Older explainers](https://yomu.fyi/visual-explainers/page/5.md)