Skills

Ship AI Products in Production

4 milestones · 12 weeks

Ship a real AI-powered product to production and keep it live with real users — proof is usage data and a technical post-mortem, not a tutorial.

Build an AI Agent from Scratch

3 milestones · 8 weeks

Design, build, and deploy an AI agent that takes actions in the world autonomously.

Ship an AI Product to Real Users

4 milestones · 10 weeks

Take an AI feature or product from working demo to 10 active users outside your network. Proven by real usage data, not by the code shipping.

Drug Interaction Analysis

0 milestones · 6 weeks

Analyse a real polypharmacy case for clinically significant drug interactions. Proof requires: (a) a documented case (real or realistic) with the patient's medication list (minimum 5 concurrent medications); (b) identification of all clinically significant interactions using a recognised interaction checker (Stockley's, BNF, Drugs.com interaction checker) with the mechanism of each interaction documented; and (c) a recommended management plan for each interaction identified. Reviewed by a registered pharmacist. The proof is the analysis and reasoning — not AI-generated output submitted as evidence. The pharmacist confirms the interaction assessment is accurate before the proof is considered complete.

Applied Ethics Case Analysis

0 milestones · 8 weeks

Analyse a real ethical dilemma — bioethics, AI ethics, business ethics, environmental ethics, or political philosophy — using at least two competing ethical frameworks (consequentialism, deontology, virtue ethics, contractualism, or equivalent). Name each framework, cite primary or secondary philosophical sources, and demonstrate what each framework concludes about the dilemma and why they conflict. Proof is the written analysis plus a documented review by a philosophy academic or applied ethics practitioner who challenges your framework application — specifically, whether each framework is being applied correctly and whether you have identified the genuine point of ethical conflict. Their written questions and your written responses are required parts of the proof.

Environmental Monitoring Data Collection

0 milestones · 6 weeks

Collect real environmental monitoring data from a specific named location using a documented methodology: air quality (PM2.5, NO₂, CO₂), water quality (pH, dissolved oxygen, turbidity, conductivity), biodiversity count, temperature, or equivalent environmental parameter. Document location coordinates, measurement dates and times, equipment or method used, and any data quality flags. For students without field access or equipment, monitoring data from named open government or citizen science networks is accepted: NOAA monitoring stations, EPA Air Quality System (free), OpenAQ (free), iNaturalist (for biodiversity), or Copernicus atmospheric data — but the provenance (station name, coordinates, data version) must be fully documented. The proof is a real documented dataset with full provenance — not a description of what you would measure, but actual data from an actual named location. Reviewed by an environmental scientist who examines the methodology for sampling bias and asks how the monitoring protocol would need to change to answer a specific environmental question.

Climate and Environmental Open Data Analysis

0 milestones · 8 weeks

Analyse a real named publicly available climate or environmental dataset: NOAA Global Surface Temperature, NASA GISS Surface Temperature Analysis, Copernicus Climate Change Service ERA5 reanalysis, EPA Air Quality System historical data, or equivalent. The proof is a documented analysis report: data source and version, preprocessing decisions with rationale, statistical analysis (trend detection, anomaly identification, spatial or temporal comparison), uncertainty quantification, and written conclusions with specific physical interpretation. All sources listed are free and publicly accessible. Real climate datasets contain real measurement uncertainty, coverage gaps, and confounds that require genuine scientific reasoning to navigate — rigorous analysis produces real artifacts of scientific value. Reviewed by an environmental scientist who challenges the statistical methodology — specifically asking whether the identified trend is robust to a specific alternative specification — and asks you to interpret a specific dataset feature you did not discuss in your report.

Environmental Policy Brief

0 milestones · 8 weeks

Write a policy brief (maximum 4 pages) on a specific real environmental issue for a specific named decision-maker — a government environment agency, a planning authority, a local authority, a conservation body, or a named NGO. The evidence base must draw primarily on environmental monitoring data and ecological science: air quality records, biodiversity survey data, climate projections, EIA findings, or conservation science research. The brief must not rely solely on social science or health evidence — those are covered by separate outcomes (social-policy-analysis-recommendation and health-policy-brief respectively). What distinguishes this outcome is its evidence domain: environmental science data analysed through an environmental regulatory and conservation framework, for a decision-maker whose remit is environmental protection or resource management. The policy recommendation must be specific and actionable. Reviewed by an environmental scientist or environmental policy practitioner who challenges the evidence base — specifically, whether the cited monitoring data supports the claimed environmental impact at the stated confidence level — and asks what monitoring data would change the recommendation.

Air Pollution Control Analysis and Sizing

0 milestones · 6 weeks

Select and size an air pollution control technology for a defined emission source and pollutant, documenting the analysis from source characterisation to technology recommendation. The analysis must include: a source characterisation (pollutant species, flowrate, concentration, temperature, and relevant physical properties), a screening of at least 3 control technologies (with documented performance data — collection efficiency, operating temperature range, pressure drop) against the source characteristics and applicable emission limit (regulatory limit from a named standard), a sizing calculation for the selected technology (device dimensions or operating parameters with documented methodology), a performance prediction showing the expected outlet concentration versus the emission limit, and a capital and operating cost comparison (order of magnitude, from published EPA cost data or equivalent). Preferred proof: analysis for a real emission source. Accessible alternative: EPA AP-42 (free, online compilation of emission factors and control technology performance data), EPA Air Pollution Control Cost Manual (free, online) applied to a publicly documented industrial source — no measurement equipment required, analysis uses published data. Proof artifacts: the source characterisation and technology comparison (analysis artifact) and the sizing calculation and performance prediction (documentation artifact). Verification: an environmental engineer reviews the performance prediction — 'your calculated control efficiency is X%; what would the outlet concentration be during a process upset when flow increases by 20%?' — requiring specific reasoning from your own sizing.

Visual Identity System

0 milestones · 8 weeks

Design a complete visual identity system responding to a real or set brief. Triad: concept/ideation artifact (brief response document and moodboard documenting the strategic rationale — why this mark, this palette, this typeface for this brief), technical execution artifact (production-ready logo system: primary mark, secondary variants, colour palette in HEX/CMYK/Pantone, and typographic pairings in vector format), critique/documentation artifact (brand guidelines document with usage rules, misuse examples, and written design rationale). Proof it is your own work: live studio critique where a working graphic designer challenges the strategic and aesthetic decisions — the critique response document showing what was challenged, what changed, and why is submitted as proof. Proof: brief document, production files (SVG/AI/PDF), brand guidelines PDF, critique notes. Verifier: practising graphic designer or graphic design tutor — no peer-only verification.

Studio Practice Development

0 milestones · 16 weeks

Documented studio practice over a minimum 12-week period developing a coherent body of work. Triad: concept/ideation artifact (artist statement and practice rationale: what artistic questions or aesthetic territories are being explored — this must make a genuine intellectual argument, not describe the work; must reference at least 3 practitioners whose work is relevant and explain why), technical execution artifact (developed body of work — minimum 5 resolved pieces in a coherent series or development sequence — documented through high-quality photographs of the finished works), documentation/critique artifact (process portfolio: material experimentation notes, iteration records, photographs of resolved and unresolved work dated across the production period — demonstrating real studio time, not a single production run). Proof it is your own work: studio crit where a fine art tutor or professional artist challenges conceptual and formal decisions in real time — an AI can produce a plausible artist statement and digitally generated images, but it cannot participate in a documented crit where specific decisions about a specific body of physical or digital work are challenged. Proof: artist statement (with referenced practitioners), 20+ dated process images, final work documentation, crit notes. Verifier: fine art tutor or professional artist — no peer-only verification.

Critical Art Research Essay

0 milestones · 10 weeks

Extended critical essay situating the student's own practice within art theory, art history, or contemporary critical discourse. Triad: concept/ideation artifact (research question and argument framework: what specific critical question this essay addresses — a genuine argument, not a survey; the question must connect directly to the student's own practice), documentation/critique artifact (critical essay of 2,000–4,000 words engaging with primary sources — artist statements, critical theory texts, exhibition catalogues, peer-reviewed art criticism — making an original argument that connects specific aspects of the student's practice to a wider critical discourse; minimum 6 primary sources cited and substantively engaged, not merely referenced in passing). Proof it is your own work: the essay must argue for a specific position about the student's own practice — an AI could produce a generic art theory essay, but an argument that connects specific claims to specific works the student has actually made requires knowledge of that practice; the studio crit where the essay's argument is challenged tests this. Studio critique: art history academic or fine art tutor conducts a documented verbal challenge to the essay's central argument. Proof: essay document, challenge-and-response notes. Verifier: art history academic or fine art tutor with postgraduate qualification in art history or critical theory.

User Research & Synthesis

0 milestones · 10 weeks

User research study conducted with minimum 6 participants and synthesised to professional standard. Triad: concept/ideation artifact (research plan submitted before data collection begins: research questions, methodology rationale — why interviews vs surveys vs observation, participant recruitment criteria and screening questions, discussion guide or observation protocol), technical execution artifact (research data collected from real participants: signed consent forms, session notes or recordings, and completed data set — the participant evidence is the unfakeable layer), documentation/critique artifact (synthesis report with documented thematic analysis: how raw data was coded, what themes emerged, how themes were prioritised, and what the design implications are — minimum 1,500 words of analysis). Proof it is your own work: participant consent forms and session records cannot be fabricated without real people; the synthesis process — specific themes emerging from specific quotes from specific participants — traces back to real sessions in a way that AI-generated data cannot. Studio critique: UX researcher or HCI practitioner challenges the synthesis methodology and the validity of the design implications drawn. Proof: research plan, consent forms, session records, synthesis report, critique notes. Verifier: UX researcher, HCI researcher, or senior UX practitioner with research methodology experience — no peer-only verification.