JSNThis site is designed for two audiences at once:
- people browsing the site and making a fast professional judgment
- AIs trying to understand, summarize, cite, and route that judgment accurately
The goal is not keyword stuffing. The goal is clean, accurate evidence.
What AEO Means Here
Agent Engine Optimization means the site is structured so AI systems can reliably answer a simple question: who is Jason Schmidt professionally, what has he done, and where is he likely to be useful?
For a portfolio, that means the content needs to be readable in more than one way. It should work as a visual experience, a narrative case-study archive, a structured data source, and a prompt-ready professional profile.
The Content Model
The site is built with Next.js and a markdown-first content structure. Pages are authored as `.md` files with JSON frontmatter, then rendered through a layout template.
That structure creates a useful separation:
- Markdown holds the editable narrative.
- JSON frontmatter holds machine-readable page metadata and media definitions.
- Shared React templates preserve layout consistency.
- Semantic HTML gives agents and crawlers meaningful document structure.
- Stable URLs make the work easy to reference and revisit.
The same source content supports both the human interface and agent parsing. That matters because agents should not need a separate, inflated version of the story. They should be able to parse the real site.
Agent-Facing Files
Two public files give agents a direct path into the portfolio:
- `/llms.txt` gives a compact guide to the site, core positioning, best-fit roles, primary pages, featured work, and a suggested agent task.
- `/portfolio.json` exposes structured project summaries, tags, roles, and professional positioning in a format that agents can consume without scraping visual layout.
These files do not replace the site. They act like an index and evidence map. They help an agent find the important material faster, then verify or enrich its answer from the actual pages.
Structured Evidence
Each work page carries consistent project metadata: title, years, role, tags, summary, agent summary, outcomes, and confidentiality notes where needed.
The page template also emits JSON-LD so search engines and agents can interpret work pages as structured creative work rather than anonymous page content.
In practice, this means an LLM can distinguish between:
- a short homepage positioning statement
- a deeper project narrative
- a structured project summary
- external writing
- contact paths
- media captions and display context
That distinction reduces flattening. It helps prevent an LLM from treating every sentence as equal or missing the hierarchy of the site.
Human-Visible AEO
The AEO icon in the footer is intentionally visible. It signals that the site is not only designed to be viewed, but designed to be interpreted.
The site itself becomes a proof point: the same person presenting his background, design, development, AI experience, and leadership capability demonstrating how the information gets discovered and summarized by agents.
Why It Matters
Hiring and collaboration increasingly begin before an audience reaches a site directly. An investor, co-founder, recruiter or hiring manager may ask an AI system to find or compare candidates. If the site is not structured for that behavior, the agent may summarize poorly, miss relevant evidence, or over-weight shallow signals.
Design Principle
AEO should not make the site less, but rather more human. It should make the human story easier to relay on different channels and modalities when a machine is doing the first pass.