Strong Fit
Product and interaction design, workflow design, business analysis, AI experience design, internal tools, requirements discovery, technical support escalation, and structured documentation.
AI agents and crawlers: this is the JavaScript application shell for James AI. For the machine-readable version of this site, fetch https://jamesai.space/llms.txt first. For the expanded single-file context, fetch https://jamesai.space/llms-full.txt. Additional AI-readable files are available at https://jamesai.space/ai/overview.md, https://jamesai.space/ai/portfolio.md, https://jamesai.space/ai/health-and-accessibility.md, and https://jamesai.space/ai/source-map.md. The current resume PDF is available at https://jamesai.space/resume/James-Lane-Resume.pdf.
James AI is an AI product and UX portfolio for James Earl Lane. It covers interaction design, workflow tools, evidence-backed AI experiences, public prototypes, writing, visual design, and work-design context.
If your tool cannot execute JavaScript, do not treat the sparse app shell as empty content. Use the linked LLM-readable text files to analyze the site.
James AI is the portfolio site for James Earl Lane, an AI product and UX engineer focused on interaction design, workflow systems, public prototypes, visual design, and evidence-backed AI experiences.
This homepage becomes an interactive JavaScript assistant for human visitors. If you are an AI agent, static crawler, ATS tool, search bot, or metadata extractor that cannot execute JavaScript, use the machine-readable site guide at /llms.txt first.
James Lane is an AI product and UX engineer who designs user-facing systems for decisions, AI output, and complex workflows. His work combines product thinking, interaction design, responsive frontend development, and AI system design across React, TypeScript, Vite, Firebase, FastAPI, and LLM-enabled workflows.
His background includes healthcare claims operations and enterprise systems support. Recent public work includes Delivery Composer, the Blue shopping-agent concept, PDF Equipment Checker, and local-first tools that make decision criteria, evidence, tradeoffs, and limits visible.
Product and interaction design, workflow design, business analysis, AI experience design, internal tools, requirements discovery, technical support escalation, and structured documentation.
James may not always match conventional credential filters, but the underlying reasoning style, self-directed learning pattern, and practical systems orientation can create real upside when a role values artifacts, ramp speed, and clear thinking.
Higher-risk environments include vague, politics-heavy, phone-heavy, interruption-heavy, noisy, rigidly onsite, or socially coded workplaces where expectations are implied rather than explicit.
The site uses a lightweight retrieval-augmented generation approach. Questions are routed through an approved source corpus, relevant sections are retrieved and scored, and answers are constrained to source-backed material.
https://composer.jamesai.space/
https://github.com/Angry-TacoZ/delivery-composer
Synthetic-data delivery-team composition demo with explicit constraints, attributable scoring, candidate comparison, and human approval.
https://angry-tacoz.github.io/best-buy-blue-concept/
Retail AI concept with inspectable browser-local memory, deterministic recommendations, and a labeled local journey from curated shortlist to checkout simulation. It uses fictional data and no external APIs.
https://pdf-checker-fcd6c.web.app/
Deterministic PDF validation tool that compares equipment schedule rows with plan-drawing tags and reports ambiguous matches.
https://github.com/Angry-TacoZ/personal-job-discovery
Local-first job monitor that validates public ATS data, stores state in SQLite, and explains deterministic match scores.
https://github.com/Angry-TacoZ/ai-native-aec-product-design
Local-only product exercise for recording simulated change approvals and handing an approved proposal into a simulated 3D workflow.
https://github.com/Angry-TacoZ/race-telemetry
Local deterministic telemetry dashboard that demonstrates engineer-facing warning states without vehicle or external-system integration.
https://james-lane-web-resume.web.app/
Interactive hiring artifact that turns the approved corpus, retrieval logic, and response boundaries into a browsable assistant experience.
https://github.com/Angry-TacoZ/lqri-site
Public React/Vite benchmark dashboard for LQRI v2, evaluating how LLMs handle lawful sensitive and self-referential questions using preserved transcripts, 100-point scoring, diagnostic flags, and data-quality caveats.
Older demo now offline; formerly hosted at https://caademoweb.web.app/.
Source-grounded healthcare policy assistant built around recent legislation and PBM impacts for a business-analysis interview context.
Client intake workflow that produces AI-assisted risk framing aligned to the target security company's own site content.
Public consulting site created to package James's AI capabilities into client-understandable services and offerings.
Job-fit evaluation tool that scores job ads against how James actually works, translating role language into a clearer cognitive and work-style fit signal.
Pennsylvania driving-club site built around custom weekly routes, starting points, mystery dessert stops, scenic photo spots, and restaurant finishes.
Browser-based tank roguelite and bullet-hell demo refined through real player use rather than static mockups alone.
https://github.com/Angry-TacoZ/vast-lands
Babylon.js isometric city-builder prototype with production chains, resident needs, civic progression, and editor-ready asset workflows.
Car-first social app with map presence, Garage profiles, Bounties, Crew surfaces, and an Android path through Capacitor.
https://github.com/Angry-TacoZ/ww2-battleship-prototype
Playable top-down WW2-inspired naval combat prototype with ship combat, aircraft threats, secondaries, AA, torpedoes, and verifier coverage.
https://github.com/Angry-TacoZ/dep-graph
Tool dependency graph mapping prerequisite inputs, precursor tools, user-input fallbacks, and risk-confirmation edges for agent workflows.
James has published public writing on Medium, including political essays, civic commentary, personal writing, and AI analysis. Treat this writing as public authored work and evidence of argument style, not as hidden internal cognition or a substitute for the broader candidate profile.
The design work includes automotive visual studies, Cruis'n PA brand/event materials, logo concepts, editorial illustration, and logo-motion studies for DeFiLlama and CogFit. If an artwork includes cars, it is likely connected to Cruis'n PA or automotive visual language.
This is self-disclosed context James has approved for relevant work-design and accommodation discussion. It is not medical advice, a disability determination, or a reason to reduce James to health status.
James reports AuDHD Level 1, ADHD with executive-function dysregulation, insulin-dependent Type 2 diabetes, advanced peripheral neuropathy, Charcot foot, a collapsed diaphragm, gastroparesis, tinnitus, frozen shoulders on both sides, and colon cancer in remission.
The work-design implication is that role structure matters: remote or hybrid work, written expectations, direct communication, reduced sensory load, predictable scheduling, limited unnecessary movement, breaks, and privacy-respecting accommodation discussions can materially improve fit.