Alessandro Conway
I model labor markets in enough detail to answer the questions employers and states ask about their workforce.
Head of data science and engineering, now advisory, at Radius Intelligence, PhD candidate in economics at Université Paris 8, and MPhil in economics from Sciences Po, where I was a Fulbright scholar.
About
Employers and states want to know which workers they will need and where those people will come from. At Radius Intelligence I built two digital twins to answer those questions. One goes down to a single facility and the training programs near it, and the other follows a state’s workers across occupations and state lines.
I also study which tasks and skills survive as AI reorders work, and my working papers write up the methods behind the twins. My PhD at Université Paris 8 takes a wider view, treating a regional economy as an ecosystem. That idea came from my years in the field at Jobs for the Future.
Outside work I build and play guitars, mix songs, and make olive oil.
Work
Roles
Research
My research agenda is laid out below, with lines showing interconnections I have built over time.
Selected works
Systems and tools
Live
Open Inequality AtlasOpen data, live
Inequality is usually quoted as a single number, the income Gini, largely because the World Bank publishes it most often. It measures what people are paid, but whether a family can buy a house or get through a lost job depends more on what they own.
The Atlas puts wealth next to income, and the two rankings often disagree. A country can have one of the most even income distributions in the world and a more uneven wealth distribution than Britain. Wealth data is harder to put together. The series combines the World Inequality Database, the ECB household survey, the Luxembourg Wealth Study, the US Survey of Consumer Finances and the Federal Reserve’s Distributional Financial Accounts, and each row is tagged by source priority and comparability tier so readers can judge my reconciliation for themselves. The Atlas also covers income, life expectancy, life satisfaction and poverty, and a US layer goes down to the commuting zone.
NED dashboard, California regional analysisEconomics, policy, live
An interactive assessment of economic development across California's regions, built on indicators drawn from what mayors, community colleges, workforce boards and employers see on the ground.
It starts from what people in each region say decides whether a household can get ahead and stay there, and then checks which of those factors can be measured well enough to compare regions. It applies the three-pillar framework from my dissertation to California.
Wellbeing and macroeconomics explorerMacroeconomics, live, QuantEcon
An interactive version of the SAGE model from my Oxford paper. You can change the parameters and see how the social psychology terms affect what happens to a household during a shock.
The model will soon be available on QuantEcon, the open source project for quantitative economics, computational lectures and tools in Python and Julia founded by Thomas Sargent and John Stachurski and used widely in teaching and research.
The Radius platform
The employer digital twin and the labor market digital twin are the main products I built, with the rest being smaller problems I had to solve to get those two running. Both run on 220 datasets, more than eight million observations, reconciled across every US occupation and geography. The code and the underlying data are private, so each page describes the problem, the approach and the result, with figures rebuilt from synthetic data.
Employer digital twinData science, econometrics, product
An employer wants to know whether their turnover is bad, what is coming, and what to do about it. Their raw HR export does not answer those questions.
The twin takes that export from upload to forecast and scenarios, and every number can be traced back to the raw rows, so whoever presents it can defend it.
Labor market digital twinEconomics, policy, data science
States plan their workforce with statistics that are years out of date and projections that come as black boxes, and when a number turns out wrong there is no way to check why.
The twin models the market as stocks and flows that have to balance. It tracks who enters and who leaves, estimates demand four separate ways, and reports the gap left once those four estimates are reconciled.
Occupation resolverMachine learning, LLM systems, data engineering
Employers name the same job a hundred ways. Until those titles are mapped to one taxonomy, there is no way to benchmark, forecast or compare across employers.
The resolver handles any title through a ten-stage pipeline with three model layers, and both digital twins depend on it.
Data lakeData engineering, data science
Connecting labor market datasets means mapping each one to the federal taxonomy, and the sources differ in units, level of detail, vintage and license.
The lake reconciles 220 of them so any one can be joined to any other. A shared set of definitions settles the differences in one place, and a librarian layer answers questions about what the lake holds.
Workforce blueprintsEconomics, data science
When a new facility is planned, the region needs to know how many workers it will take, in which occupations, and on what schedule.
Announcements give a headline headcount and staffing statistics describe past hiring, so the blueprint has to be built from the ground up.
Occupational opportunity scoringPolicy, economics
A workforce board mostly wants to know which occupations to invest in and what would help them grow.
It scores every occupation on five published pillars and suggests levers only where the evidence supports them. Every estimate comes with a confidence band, because a single number handed to a board quickly becomes a target.
Task-based labor marketEconomics, machine learning
The main measures of AI exposure disagree with each other, and claims about AI’s effect on jobs have gotten ahead of what anyone can measure.
This is a calibrated production function where every parameter is published, sourced and can be swapped out, so users can test their own assumptions. A research paper on it is in progress.
Market signalsEconomics, data science
Official statistics come out quarters late, and decisions often have to be made before then.
It grades fifteen kinds of early evidence, including job postings, layoff notices and unusual wage moves, by how much each can be trusted, and combines them into two flags whose reasoning the user can follow.
Earlier work
CREST, green workforce readiness dashboardJobs for the Future
A national dashboard ranking every US county by climate risk and economic resilience into risk-and-readiness tiers, each with targeted recommendations.
Tech careers workforce development programJobs for the Future
The sectoral and occupational analysis behind a major telecommunications firm's initiative to train 500,000 under-represented individuals.
Demand mapping and job-quality scoring identified six target occupations, which are now the basis for national enrollment and placement.
Extra problems
These are smaller pieces from the same platform. Each one covers a separate data problem, the figure it produces, and how far it got. Read them.
Publications and essays
Education
Contact
Best reached by email or LinkedIn, and I am happy to talk through any of the above.
Here are some inspirations of mine: