Start here
How AI Works
Published 2026-02-06
A plain-English explanation of what happens when a user sends a prompt and receives a response.
Open guideLearn
Plain-English technical guides for software and AI decisions.
These pages are for business owners, operators, and software teams that need enough technical understanding to make better decisions. The goal is not jargon. The goal is clear ownership, safer AI workflows, and fewer expensive surprises.
Recommended Path
Start with the simple explanation, then go deeper only where it helps.
Deeper technical view
AI Pipeline Breakdown
Published 2026-02-20
A full walkthrough of the AI pipeline from request intake, tokens, attention, sampling, safety checks, and streaming.
Open guideModel strategy
AI Model Size Guide
Published 2026-03-21
A practical guide to large models, small models, deployment constraints, training tradeoffs, and hybrid AI systems.
Open guideModel comparison
ChatGPT vs Gemini vs Claude
Published 2026-04-03
A practical guide to why the major AI assistants behave differently across architecture, context, alignment, and reasoning.
Open guideAgent systems
The Agent Loop
Published 2026-04-18
How AI systems go from answering questions to planning, using tools, observing results, and completing real work.
Open guideProduction agents
AI Agent Best Practices
Published 2026-05-02
How to scope, design, guard, evaluate, monitor, and maintain AI agents that call tools and affect real workflows.
Open guideComputer vision
iOS Vision vs Android Computer Vision
Published 2026-05-29
A practical comparison of Apple's Vision Framework, Core ML, TensorFlow Lite, ML Kit, and heuristic mobile vision pipelines.
Open guideMore Resources
Use these when the decision moves from curiosity to implementation.
The education pages support the main HerbDev offer. They help a visitor understand the work before asking for rescue, AI consulting, or maintenance ownership.
Agent Systems
AI Agent Patterns
Published 2026-01-09
Common agent architectures for intake, retrieval, planning, scheduling, decision support, and human approvals.
Agent Loop Optimization
Published 2026-06-27
How harness, API, and inference layers reduce repeated work in production agent loops.
AI-Native Engineering
Published 2026-03-07
How AI-assisted engineering changes context, specification, verification, team structure, and ownership.
Model Strategy
GPUs Just Lost the Token Race
Published 2026-08-09
Why Groq, Cerebras, and SambaNova are changing AI inference with the LPU, wafer-scale compute, and reconfigurable dataflow.
How Big Models Teach Small Models
Published 2026-07-25
Knowledge distillation, teacher models, student models, soft labels, and deployable AI.
LLM Memory Cost
Published 2026-07-12
Why KV cache, long context, batch size, and decoding make model memory expensive.
LLMs in Search Architecture
Published 2026-06-13
Where LLMs fit in production search: catalog enrichment, query understanding, embeddings, and guardrails.
Mobile and Business Architecture
Embedded AI vs Cloud AI
Published 2026-05-16
How to choose mobile AI architecture for cost, speed, privacy, offline use, and long-term maintenance.
iOS Vision vs Android Computer Vision
Published 2026-05-29
Apple Vision Framework, Core ML, TensorFlow Lite, ML Kit, and heuristic mobile vision pipelines.
App Revenue Calculator
Published 2026-01-23
Estimate subscription app revenue across subscriber counts, monthly pricing, yearly pricing, and platform fees.
HerbDev Perspective
Learning should make the next decision easier.
If a project is already stalled, unstable, or difficult to release, education is useful only when it leads to clearer ownership. The fastest path is usually a technical triage, a short list of blockers, and a practical rescue plan.