Applied AI and technical operations work for teams that need clearer processes, human review checkpoints, and defensible handoffs.
Start with the four core technical flagships below. Employer, buyer, and research paths remain separate because they answer different questions; none implies customer, revenue, or compliance claims.
These are the primary technical projects. Their ordering is a review sequence, not a claim that one project is universally stronger or more validated than another.
Local-first workspace for tracing source evidence through explicit hypotheses and decisions while keeping counterevidence and decision history inspectable.
Why review itA production Vercel application and public source demonstrate the evidence model, local-first boundary, explainable scoring, and regression-oriented verification. Practitioner adoption remains under evaluation.
Separate authoritative repositoryProvider-neutral Python reference implementation for normalizing coding-agent session records with explicit fidelity, privacy, and historical-projection boundaries.
Why review itPublished package and public source document ATIF normalization, best-effort secret redaction, fidelity accounting, and the limits of historical observability. It does not claim native session rehydration or original runtime telemetry.
Separate authoritative repository
Visual graph-based simulation lab for complex systems, bottlenecks, cascade risk, reproducible experiments, SSI scoring, optimization suggestions, and reports.
Why review itMost technically original flagship: deterministic engines, editable graph models, multi-seed analysis, local persistence, tests, architecture documentation, and a Vercel deployment.
Separate authoritative repository
Local-first visual workflow canvas for turning fragmented notes, transcripts, logs, and research inputs into structured, reviewable deliverables.
Why review itMost complete product asset: consolidated default branch, browser and unit validation, portable exports, buyer transfer materials, and consent-gated OpenAI/Gemini BYOK assistance.
Separate authoritative repositoryThe strongest external signals here are narrow and specific: merged upstream contributions and a published interoperability implementation. They do not imply customers, revenue, broad adoption, or endorsement beyond the linked evidence.
Scoped security, correctness, performance, and regression-coverage contributions were independently reviewed and merged upstream in PRs #299, #319, #320, and #322.
PR #9619 was merged upstream with a section-ordering fix that keeps visible task order aligned with existing move actions, with regression coverage across project and tag contexts.
A published Python package and technical article document ATIF normalization, fidelity accounting, redaction boundaries, and the distinction between portable history and native session resumption.
13 supporting projects
Showing 13 of 13 projects.
| Project | Status and authority | What it does and evidence | Links | Tech / framing |
|---|---|---|---|---|
 RedactReady ProPrivacy intelligence | LivePortfolio workspace | Reviews document packets locally, detects sensitive information, scores human-risk signals, maps evidence, supports redaction, and exports reports. Evidence: Strongest privacy and document-intelligence workflow: deterministic analysis, local-first boundaries, polished review UX, tests, and case-study documentation. Live Vercel production route verified. | ||
 ProcessHarborTechnical operations | LivePortfolio workspace | Turns rough operational inputs into reviewable SOPs, onboarding checklists, knowledge-base drafts, gap reports, versions, and export bundles. Evidence: Clearest role fit for technical operations, documentation, enablement, and knowledge management, with validation boundaries and a database-ready path. Live Vercel production route verified. | ||
 ScamShield AIConsumer safety | LivePortfolio workspace | Assesses suspicious messages, organizes evidence, suggests safer next steps, locates official reporting channels, and exports a structured packet. Evidence: Shows public-interest product judgment, explainable deterministic scoring, privacy-conscious workflow design, accessible guidance, and automated tests. Live Vercel production route verified. | ||
 RedactReadyPrivacy tool | LivePortfolio workspace | Reviews sensitive files locally, highlights likely private data, supports manual redaction, and exports verification-oriented results. Evidence: Demonstrates document workflows, local processing, practical threat-model boundaries, true flattened redaction output, and security-minded UX. Live Vercel production route verified. | ||
 LayerForge StudioCanvas editor | LivePortfolio workspace | Browser image editor with layered raster documents, painting tools, selections, filters, undo, persistence, and exports. Evidence: Strong frontend and interaction-depth evidence, including Canvas architecture, command history, local persistence, and dense product UI. Live Vercel production route verified. | ||
 FocusForgeProductivity game | LivePortfolio workspace | Turns focus sessions into civilization growth, research unlocks, streak history, and durable local progress. Evidence: Shows product-system thinking, stateful UX, persistent client data, testable game rules, and responsive design. Live Vercel production route verified. | ||
 VariantVision ProResearch tool | LivePortfolio workspace | Educational bioinformatics workbench for genetic-variant review, source provenance, amino-acid comparison, evidence scoring, and non-diagnostic reports. Evidence: Shows complex-domain product architecture, source-transparency UX, research support, and careful non-diagnostic scope boundaries. Live Vercel production route verified. | ||
 AstraAI chat workspace | LocalPortfolio workspace | Local AI chat workspace with a React interface, Express API layer, model configuration visibility, Markdown rendering, and transcript export. Evidence: Shows a credible provider-backed application structure beyond a thin interface, with visible configuration and missing-key states. | ||
 Nexus PlayPlatform demo | LocalPortfolio workspace | Local game storefront and library prototype with catalog browsing, cart, wishlist, simulated checkout, owned-library state, and install-queue concepts. Evidence: Demonstrates product-state modeling, platform-style consumer flows, and UI architecture without representing simulated commerce as a real marketplace. | ||
 QuoteForge LocalAgency quoting product | CommercialSeparate private source repository | White-label quote-calculator and lead-capture package for agencies, freelancers, WordPress developers, and local-service website implementers. Evidence: Ten calculator templates, embed and WordPress paths, buyer documentation, licensing, QA scripts, sales materials, and release packaging. | ||
Amino Acid WorkbenchEducation tool | LiveLegacy static artifact in this repository | Educational workbench concept for comparing amino-acid substitutions and building evidence dossiers. Evidence: A preserved legacy static build is live on Vercel; the original editable source was not available during migration. | ||
GardenGridPlanning tool | LivePortfolio workspace (recovered local source) | Garden-layout planning concept for arranging beds, companion-planting notes, and seasonal planning. Evidence: Recovered local source is preserved in this repository and deployed to Vercel after type, lint, and production-build checks. | ||
HearthLinkCommunity concept | LiveLegacy static artifact in this repository | Peer-to-peer neighborhood-hub concept with explicit offline and signaling limitations. Evidence: A preserved legacy static demo is live on Vercel. It intentionally runs in offline/demo mode without a signaling server, so no multi-peer service is claimed. |