Airfield
The photograph is a real airport. Choose a marker on the runway, taxiway, terminal, or aircraft, or open any stop from the list. If you want the work without the airport, the plain brief is the same evidence in a short read.
Aerial photograph by Curtis Cheng on Pexels
Every stop, as a list
The same destinations, without the field. Screen readers and quick visits can use this list.
Runway
- Runway 25 · Citi Project Manager (contract) at Citi, coordinating the response and explanation layer for a credit-card fraud scoring model.
- Runway 16 · Ants Corp Project Manager at Ants Corp. Software, cloud migration, cybersecurity, and standardized intake.
- Runway 09 · NaviSite Product Analyst at NaviSite. Agile delivery support, acceptance criteria, and vendor performance reporting.
Taxiway
- Taxiway D · Discover Are we solving the right problem? Customer evidence, a problem statement, and explicit assumptions.
- Taxiway P · Prioritize What deserves the next investment? Reach, impact, confidence, and effort, with uncertainty left visible.
- Taxiway E · Evaluate What evidence would justify a release? Rubrics, representative tasks, and a repeatable ship decision.
- Taxiway L · Deliver Can the team execute the plan? Owners, dependencies, and release gates a team can inspect.
Terminal
- Terminal 1 · Case studies Written from real engagements, kept at a level that protects employer and client confidentiality.
- Terminal 2 · Writing Notes on release gates, prioritization, and how to read a project.
- Terminal 3 · Project evidence Evidence packs, maturity labels, and walkthroughs for every listed project.
Gate
- Gate C1 · Citi case study Coordinating fraud-model responses at Citi.
- Gate A2 · Ants Corp case study Project intake and onboarding at Ants Corp.
- Gate W1 · LLM release gates An eval-gated release loop that catches regressions before rollout.
- Gate W2 · RICE for AI features The version of RICE used when the feature is an LLM.
- Gate W3 · Briefly teardown A design review of the workflow described for Briefly PRD AI, with open evaluation questions.
- Gate W4 · Inspectable projects A practical review guide for scope, ownership, evidence, and limitations.
Aircraft
- Autonomous Product Analyst Complete. Investigates why product metrics changed through governed definitions, safe SQL, eight diagnostic dimensions, contributor ranking, statistical evidence, and release context.
- AI Product Discovery Engine Complete. Unifies interviews, support tickets, reviews, surveys, and analytics to detect customer problems, rank opportunities, and preserve the evidence behind every recommendation.
- AB Advisor Complete. A Bayesian experimentation workspace that turns posterior probability, expected lift, credible intervals, and expected loss into an evidence-based ship decision.
- AI Experimentation Platform Complete. A working local experiment workspace with saved hypotheses, sample planning, entered conversion counts, calculated readouts, guardrail checks, review decisions, and JSON backup; rollout controls are planning state.
- Feedback Prioritizer Complete. Converts customer feedback into deduplicated themes and ranked product opportunities so teams can prioritize work with traceable evidence.
- SprintForge Complete. Runs the delivery cycle from backlog prioritization and sprint planning through Kanban execution, burndown tracking, retrospectives, and dependency visibility.
- Briefly PRD AI Live. Transforms rough product concepts into decision-ready PRDs with guided intake, user stories, acceptance criteria, success metrics, and Markdown export.
- AI Project Manager Assistant Complete. Turns an early brief and team roster into a work breakdown structure, milestones, effort estimates, resource assignments, and a RAID register.
- ProdMind In progress. An experimental workspace connecting customer evidence, prioritization, and release decisions.
- Sentinel Eval Harness Complete. An evaluation harness for AI systems with repeatable datasets, deterministic and quality scorers, mutation testing, statistical comparison, persisted runs, an API, dashboard, and CI release gate.
- Starter AI Agents Complete. A collection of 16 runnable agents spanning research, data analysis, visualization, travel, multimodal workflows, and multi-agent orchestration.
- ForgeMind In progress. An assembly-line process intelligence system with event-sourced diagnosis, multimodal perception, governed recovery, and operator confirmation; hardware and model services are required for the full demonstration.
- MCP Price Tracker Complete. A local MCP product with persistent price observations, constrained JSON-LD collection, currency-aware comparisons, source-linked history, threshold checks, and a browser workspace.
- MCP Review Intelligence Complete. A local MCP review product with validated CSV/JSON imports, deduplication, English sentiment analysis, evidence-linked aspect themes, product and period comparisons, and a browser workspace.
- FraudShield Live. A live rules-based e-commerce risk prototype showing transaction-level risk factors.
- TikTok Shop Risk Intelligence Complete. A risk-intelligence project for identifying and explaining operational and marketplace signals in social-commerce activity.
- Wine Quality Prediction Planned. A planning-stage concept for comparing wine-quality prediction models.
- Outage Watch In progress. A local outage-triage prototype that checks a sample service, matches incident signatures, and recommends responses for human review; it does not execute remediation or schedule monitoring.
- Autonomous Software Engineer Platform In progress. A governed issue-to-PR engineering system with repository intelligence, approval gates, isolated workspaces, verification evidence, evaluation interfaces, and a secure IDE-style review console.
- SecondChance Lab In progress. A deterministic chaos-testing toolkit for ambiguous user-journey outcomes, using recovery contracts, interruption injection, an effect ledger, exactly-once invariants, and CI evidence reports.
- Watermark Cleanup Live. A privacy-first, human-in-the-loop workspace for reviewing and cleaning authorized overlay-like content from images, PDFs, and text documents.
- AI Job Copilot In progress. A job-search workflow that organizes role research, application evidence, and AI-assisted preparation in one workspace.
- Resume Matcher Complete. A client-first resume tailoring app that grounds every rewrite in verified experience and exports an ATS-friendly document without storing the source resume.
- Pulse Auth Complete. An edge-native OAuth2 and OIDC identity service with email and GitHub sign-in, TOTP MFA, rotating refresh tokens with reuse detection, multi-tenant RBAC, and revocable Durable Object sessions.
- Local Dining Intelligence In progress. A local discovery product that organizes restaurant signals into practical dining recommendations and decision support.
Control tower
- Control tower · Ask Madan Questions are answered from the published profile and public GitHub repositories. If the model is unavailable, the tower says so.
Cargo
- Cargo · Templates PRD, evaluation rubric, launch checklist, and decision log. Files you can take with you.
- Cargo bay · One-page PRD A single-screen PRD for a feature that needs an owner and a shipping date.
- Cargo bay · LLM eval rubric A rubric for LLM-as-judge scoring on customer-facing features.
- Cargo bay · Launch checklist The checklist used before flipping a launch flag.
- Cargo bay · Decision log Evidence, alternatives, owner, and the trigger for reviewing a decision.
Information desk
- Information desk · Plain brief The same career evidence, written for a fast read. No airport to operate.
Ground radio
- Ground radio · Contact Email, LinkedIn, GitHub, and a message form. Seattle, WA.
Flight information
- Flight information · Now What the current week looks like, and the roles that are boarding.
Training center
- Training center · Toolkit and credentials Delivery methods, tools, and the degrees and credentials behind the work.