51.5074° N0.1278° W

00 — The Thesis

Every question had to be made answerable.

Fragmented evidence, institutional records that disagree, boundaries no one had settled. The route to an answer had to be built first — from one borough to an international market.

Three recurring areas: making a subject investigable, investigating where no route exists, keeping the evidence usable.

  • 01Evidence systems

  • 02Applied investigations and strategy

  • 03Research instruments

Scroll — or tap the beats ↓

Project 01LOCAL

What do public records show about funding linked to Barking?

01The question

Published award records and charity accounts describe different parts of B&D’s charitable economy. Read together, they show which organisations appear in each register, which published funders are visible, how recorded awards enter the borough-connected population and how reported income is distributed. They do not provide a complete measure of funding or B&D-local spending.

02Follow the money

The audit defines an 808-organisation seed population and maps each published GrantNav connection by one step. It separates awards entering from upstream non-seed funders from limited seed-to-seed and seed-to-non-seed activity, while Charity Commission accounts provide context on organisational scale and financial review signals.

03Limited circulation

The retained network is overwhelmingly inbound: £91.2m of £92.2m in observed award value enters the seed population from upstream non-seed funders. Published seed-to-seed awards total £263k and originate from the sole seed organisation observed distributing awards. From 2021 to 2024, the top tenth of reporting organisations account for 85–88% of reported annual income.

FIG. 01

Observed funding flows

Project 02GLOBAL

Do the workshops change academic careers?

01The question

The workshops were intended to strengthen international academic ties and engagement. The motivating question is whether participation changed academic careers. The available evidence answers something narrower: what patterns appear around first participation among the people whose publication records can be linked and followed?

02The cohort

Link the programme record to ORCID, OpenAlex and Crossref; resolve scholarly identities; group versions of the same output into publication families; compare the three years before and after first participation; estimate person- and calendar-year-adjusted differences; and examine co-authorship only where both workshop peers have accepted OpenAlex identities.

03The trajectory

Among 44 participants with complete seven-year observation windows, publication trajectories varied and raw output fell on average. After adjustment, the average estimate was close to zero but imprecise, so no common workshop-timed shift was detected. Of 1,138 workshop-peer pairs, 18 had accepted OpenAlex identities for both people; two first appeared as co-authors after the workshop. These are observed associations, not causal effects.

FIG. 02

The programme collaboration network

Project 03URBAN

Is access to culture equitable?

No single register describes the organisations through which London’s cultural life is produced, and access to what they offer is uneven. Codex identifies and connects them; Strata asks what becomes visible when they are read against the spatial structure of the city.

> codex map_

03A

Strata · choice network

03B
Project 04GLOBAL

Can an expansion strategy be justified rather than intuited?

01The brief

The global art market concentrates almost three quarters of its sales in three national markets — the UK, the US and China. For a consultancy already operating in one of them, expansion is not a question of ambition but of address: which markets, entered where, justified by what?

02The pipeline

Three geographies, three postures, three instruments. Planning data turns the UK’s development pipeline into a lead engine; a census-driven index funnel narrows fifty US states to five entry cities; a gateway thesis routes the China strand through two culturally-mediating localities, their identities withheld. The strands converge in an autonomous competitor sweep — twenty-four agentic analyses across eight cities and three segments.

03The shortlist

Eight cities, a register of roughly two hundred competitors, and a strategy the client can re-run as the data moves. The recommendations themselves were bought and stay withheld — what is published is the architecture that produced them, redactions included.

FIG. 04

Fifty states to five entry cities

Project 05NATIONAL

What environmental traces become visible when political power changes?

01The question

A council changes hands; a commitment is declared; something is or is not commissioned and built; and only then might an environmental signal move — while organised public action runs alongside, sometimes ahead of the politics, sometimes behind. No one system tests that chain: a commitment is not delivery, and delivery is not an effect. So five are assembled and read together. A research framework, not a proven causal chain.

02Five systems

The observatory is being assembled one system at a time: the electoral record, policy and procurement registers, ecological observation and climate services, and the record of organised public action. Each has to be obtained, made internally consistent, and placed on a shared timeline and a shared map before any relationship between two of them can be tested.

03What registers

None yet — deliberately. The instrument is being calibrated: some sources have returned structured samples, others have confirmed routes awaiting acquisition, and the geographies do not yet align. Findings appear here only when a versioned analytical release has tested a relationship.

FIG. 05

Five systems — every link untested

Methodological demonstrations, not a catalogue of commissions.

Some are self-directed, and some remain in active development. Each shows how the work is constructed and where its evidence stops.