Developer API documentation

A small, public JSON view of my selected professional experience, skills, publications, projects, and profiles. It uses the JSON Resume schema and includes links to published work. For a readable version of the same data, see my resume.

Read the mini-resume · Version 1 index · OpenAPI 3.1 schema · Agent guidance

For a short starting point and a guide to the available resources, see developer resources.

When to use it

Use the API to answer questions about my current role, find published articles, link to my public profiles, or summarize my game projects. This is a selected public overview, so it does not include my full career history, private contact details, or unpublished work. For full career history, see my LinkedIn profile. For game platforms and store links, use the Zorse game catalog API.

Quickstart

No account, API key, payment, SDK, or setup is required. Try the real endpoint directly; it is read-only and there is no separate sandbox.

curl --fail https://austingreco.com/api/v1/resume

With native JavaScript fetch in Node.js, or in a browser on this site:

const response = await fetch('https://austingreco.com/api/v1/resume');
if (!response.ok) throw new Error(`HTTP ${response.status}`);
const resume = await response.json();
console.log(resume.basics.name, resume.basics.label);

Request and response

GET /api/v1/resume returns the public mini-resume. GET /api/v1 returns a discovery index with the version’s support status, endpoint, documentation, and OpenAPI links. Both return HTTP 200 with Content-Type: application/json; charset=utf-8. HEAD returns the resource headers without a body.

Every Accept header receives JSON at these endpoints. Query parameters are ignored; there are no filters, pagination, writes, or webhooks. Conditional requests may receive HTTP 304 when the published resource has not changed. The existing /api/resume remains available as a compatibility endpoint with the same public mini-resume response.

The version index contains:

{
  "version": "v1",
  "status": "supported",
  "documentation": "https://austingreco.com/api",
  "openapi": "https://austingreco.com/openapi.json",
  "endpoints": {
    "resume": "https://austingreco.com/api/v1/resume"
  }
}

The mini-resume contains:

  • $schema identifies the pinned JSON Resume schema.
  • basics contains my name, current title, site URL, summary, public city, and professional profiles.
  • work contains selected current and earlier experience, with employer, position, URL, and summary. Entries may also include contribution highlights and role-specific keywords.
  • skills groups experience across projects into named keyword lists.
  • publications contains selected articles with a title, publisher, publication date, URL, and summary.
  • projects remains a flat list with a name, description, and URL for every entry. The optional JSON Resume entity field identifies affiliations: Bunny Hop, Tap Bat, and Yodoku use "Zorse", Austin’s independent publishing identity. The readable resume groups these games beneath Zorse and shows experiments afterward. Entries are in presentation order; use names or URLs to identify projects rather than array positions.
{
  "$schema": "https://unpkg.com/@jsonresume/schema@1.7.0/schema.json",
  "basics": {
    "name": "Austin Greco",
    "label": "Staff Software Engineer",
    "url": "https://austingreco.com/",
    "summary": "Staff Software Engineer focused on frontend architecture, React and TypeScript interfaces, and applied AI. Led UI architecture for Uber’s machine-learning tools, turning complex workflows into usable products and reusable UI systems. Independent game maker behind Zorse.",
    "location": {
      "city": "San Francisco",
      "region": "CA",
      "countryCode": "US"
    },
    "profiles": [
      {
        "network": "GitHub",
        "username": "austingreco",
        "url": "https://github.com/austingreco"
      },
      {
        "network": "LinkedIn",
        "username": "austingreco",
        "url": "https://www.linkedin.com/in/austingreco"
      }
    ]
  },
  "work": [
    {
      "name": "Uber",
      "position": "Staff Software Engineer",
      "url": "https://www.uber.com/",
      "summary": "Led frontend architecture and developer experience for Michelangelo, Uber’s machine-learning platform, and built interfaces for generative AI products.",
      "highlights": [
        "Led Michelangelo’s UI development, improving reliability through architecture rewrites and testing while supporting complex machine-learning workflows.",
        "Unified machine-learning applications on shared UI architecture with team leads and designers. Built configuration-driven forms that let engineers add workflows without specialized frontend knowledge.",
        "Led initial UI design and technical strategy for a self-service GenAI application builder. Led UI development for agent workflows and built an interface for comparing LLM behavior across configurations.",
        "Developed the Legal Redlining Agent’s React and TypeScript interface and streaming document Q&A. Led feedback sessions and feature prioritization with lawyers, and worked around Office.js constraints for responsive contract review, tracked changes, and document editing.",
        "Mentored and onboarded engineers, and presented UI architecture and AI product work to engineering leadership."
      ],
      "keywords": [
        "Frontend architecture",
        "React",
        "TypeScript",
        "Reusable components",
        "Applied AI",
        "LLM workflows",
        "LLM comparison",
        "Contract redlining",
        "Tracked changes",
        "Office.js"
      ]
    },
    {
      "name": "Intuit",
      "position": "Senior Front-End Engineer",
      "url": "https://www.intuit.com/",
      "summary": "Built React interfaces for QuickBooks Online and Demandforce."
    },
    {
      "name": "Union Pacific Railroad",
      "position": "Project Engineer",
      "url": "https://www.up.com/",
      "summary": "Led a UX and frontend team, built a responsive component library, and modernized web application standards with AngularJS."
    }
  ],
  "skills": [
    {
      "name": "Frontend and UI systems",
      "keywords": [
        "React",
        "TypeScript",
        "JavaScript",
        "HTML/CSS",
        "Frontend architecture",
        "Product interfaces",
        "Reusable components",
        "Design systems",
        "UI performance",
        "Testing"
      ]
    },
    {
      "name": "Applied AI",
      "keywords": [
        "LLM workflows",
        "LLM comparison",
        "Streaming interfaces",
        "Server-Sent Events (SSE)",
        "FastAPI",
        "LangGraph"
      ]
    },
    {
      "name": "Interactive graphics",
      "keywords": [
        "Canvas",
        "WebGL",
        "Animation",
        "Reduced-motion support"
      ]
    },
    {
      "name": "Document workflows",
      "keywords": [
        "Contract redlining",
        "Contract review",
        "Document editing",
        "Tracked changes",
        "Microsoft Word add-ins",
        "Office.js",
        "OOXML",
        "ProseMirror",
        "Tiptap",
        "Slate"
      ]
    }
  ],
  "publications": [
    {
      "name": "Scaling AI in Legal: Building Uber's Redlining Agent",
      "publisher": "Uber Engineering",
      "releaseDate": "2026-10-08",
      "url": "https://www.uber.com/us/en/blog/building-ubers-redlining-agent/",
      "summary": "Primary author of an Uber Engineering article about the design and implementation of the Legal Redlining Agent."
    }
  ],
  "projects": [
    {
      "name": "Zorse",
      "description": "Independently designed, built and published mobile games with millions of lifetime downloads.",
      "url": "https://zorse.io/"
    },
    {
      "name": "Bunny Hop",
      "description": "A pixel platformer with charge-and-release jumps, combos, and a growing cast of characters.",
      "url": "https://zorse.io/games/bunny-hop/",
      "entity": "Zorse"
    },
    {
      "name": "Tap Bat",
      "description": "A Halloween arcade game with one-touch flight through a spooky pixel cave.",
      "url": "https://zorse.io/games/tap-bat/",
      "entity": "Zorse"
    },
    {
      "name": "Yodoku",
      "description": "A calm, offline logic game with 9,999 lantern puzzles and no guessing.",
      "url": "https://zorse.io/games/yodoku/",
      "entity": "Zorse"
    },
    {
      "name": "Interactive web experiments",
      "description": "React and TypeScript playgrounds with Canvas and WebGL graphics, animation, shareable controls, and reduced-motion support.",
      "url": "https://austingreco.com/experiments"
    }
  ]
}

Errors and recovery

GET and HEAD requests to unknown /api/* paths return HTTP 404 with Content-Type: application/problem+json; charset=utf-8 and Cache-Control: no-store. These responses use RFC 9457 problem details. GET includes the following error body; HEAD has the same status and headers without a body.

type identifies the problem, title and detail explain it, status matches the HTTP status, and instance is the requested path without its query string. The stable error code, message, resolution, and documentation links remain available as extension fields. Use the code and recovery hint instead of parsing human-readable text.

{
  "type": "about:blank",
  "title": "Not Found",
  "status": 404,
  "detail": "Unknown API route. GET /api/v1/resume returns the public mini-resume.",
  "instance": "/api/v1/missing",
  "error": "api_route_not_found",
  "message": "Unknown API route. GET /api/v1/resume returns the public mini-resume.",
  "resolution": "Use GET /api/v1/resume. Read /api for examples or /openapi.json for the API schema.",
  "documentation": "https://austingreco.com/api",
  "openapi": "https://austingreco.com/openapi.json"
}

The site supports read-only access. Hosting rejects unsupported methods such as POST outside this API contract; the current deployment returns an HTML 404. Delivery or storage errors for known API resources may also return HTML. Clients should check the status and Content-Type before parsing an error. A normal missing website page keeps its HTML or negotiated Markdown 404.

Updates and retries

The resource is published with the site and may be cached by the CDN. There is no application-specific request quota or live update stream. Reuse a fetched copy where practical. GET is safe to retry after a network failure or temporary server error; honor Retry-After if the hosting service sends it. Fix an unknown route instead of retrying a 404.

Versioning and deprecation policy

API contracts use a major version in the URL. Version 1 is supported, with no retirement scheduled. New integrations should use /api/v1/resume. The existing /api/resume compatibility endpoint keeps the version 1 contract; it will not silently switch to a future major version.

Backward-compatible additions, such as optional fields or new endpoints, may be released within v1. Clients should ignore unknown response fields. Public profile values, selected work, project lists, and links may change when the site is published.

Removing or renaming fields, changing their types, adding required request parameters, or changing established behavior requires a new major version, such as /api/v2. Existing versioned URLs keep their contract and never redirect to another major version.

Before retiring a supported version or the compatibility endpoint, I provide at least 90 days’ notice on this page, a migration guide, and a working replacement throughout the notice period. Old and new endpoints remain available together during migration.

When a deprecation is announced, affected responses will include the Deprecation header and a Link header with rel="deprecation" pointing to the migration guide. Once retirement is scheduled, the Sunset header will identify the retirement date. Version 1 and the compatibility endpoint currently send neither Deprecation nor Sunset.

Discovery and related formats

The standard API discovery catalog links the version 1 API to this documentation and its OpenAPI description using RFC 9727. Use the site index for public pages, experiments, tools, and related resources.

For readable website pages, request their canonical URL with Accept: text/markdown or follow their Markdown alternate link. The mini-resume, version index, and OpenAPI description remain JSON; the API catalog uses application/linkset+json.