Miami, FL · AI software engineer · open to the right team

Joshua Lazoff

I build AI that shows its work - or says nothing.

Twenty years of production engineering for names you know - Goldman Sachs, the team behind MetaMask, Chainlink. Now I build maps of knowledge, and AI that shows its work: systems that name a real source for every answer, and stay silent when they cannot. Everything runs on hardware I own, measured by tests you can read and run yourself. I want to do this on the hardest problem you have.

The story

Twenty years, three acts

Every engineer's career is a story about what they were trusted with. Mine runs from an island where I fixed the machines myself, to the trading floor, to the front door of a new financial world. Then comes the strangest chapter: everything the career taught turned out to be preparation for a question nobody was asking yet. You will see it coming before I did. The thread underneath it has its own page.

  1. Act I · The island and the floor
  2. Origins

    San Juan · at fourteen

    The kid who fixed the machines

    Puerto Rico, before cloud anything. If you wanted a computer that worked, somebody had to build it - and by fourteen, people in San Juan were paying me to be that somebody. By high school I was also running a two-hundred-member youth organization: activities, budget, fundraising, a fifteen-person board. Machines that obey and people who volunteer are different problems. I got an early start on both. For senior year of high school, I took a one-way ticket to Florida.

  3. Engineering

    Cornell · five years

    The study of representing the world

    Information science, then a master's in engineering management. Information science asks one great question - how do you represent the real world inside a computer? It turned out to be the question I would spend my whole career answering. I taught calculus to other students and kept the books for the Hispanic engineers' chapter, because apparently I cannot be near an organization without running part of it.

    And in the summer of 2009 came my first corporate technology job: building databases at Sony Music in New York, while selling clothes in a SoHo store on the side. This site is the first public bio that job has ever appeared in.

  4. Engineering

    Three summers, one team

    Goldman Sachs, from the inside

    The same trade-processing team took me back three summers running - trust you only earn by shipping. I built the tools that check the firm's records against its own risk systems, and somewhere in there I started running the mentoring program for the incoming interns. Teaching the next class while still being one: a pattern that never stopped.

  5. Engineering

    Then the floor itself

    Systems that are not allowed to fail

    Securities technology: the systems that double-check the firm's trades and file the reports the law requires. When your software sits between a bank and its regulators, correctness is not a virtue - it is the whole job. I built a query service designed to be shared firm-wide as internal open source, and wrote in the firm's own in-house programming language. I also helped stand up a round-the-clock operation, spanning dozens of teams, that catches errors before they become anyone's worst day.

  6. The hidden hours

    Every morning, 5:50

    The double life

    Here is the part my colleagues never saw. Before the floor each day there was a study hall, an old library, and a discipline of close reading that predates every system I have ever shipped. That second life ran in parallel for years, quietly training me for work I could not yet imagine. That story has its own page.

  7. Act II · The leap, and the long apprenticeship
  8. June 2015

    Two names, one month

    I gave up the “perfect job” - the one you are not supposed to leave. That same month I incorporated two companies: LazoffTech for client engineering, and a second one for the long research project running underneath. The first has funded the second ever since. Every job that followed looked like client work and turned out, in hindsight, to be the curriculum.

  9. Research

    The next two years

    Lesson one: the data is the product

    First came a learning platform built with a team in Israel, then a ground-up rebuild of a much-loved video library. Two years of learning the lesson every senior engineer eventually learns: organizing the material is harder than writing the code - and the structure you choose for knowledge determines everything you can do with it later.

  10. September 2016

    A list called “Websites to make”

    One year in, I wrote myself a five-line plan. The first line was this website - “the main site where all the projects are listed.” It took nearly ten years, and most of that list is now real. The first line is the page you are reading.

  11. Engineering

    Six years

    The front door to a new financial world

    I joined ConsenSys through its education arm, building the platform that taught developers an entirely new way to program. I stayed six years and left as the senior engineering manager responsible for every website behind MetaMask and consensys.io - the pages millions of people walked through to enter an economy that had not existed a decade earlier. Running the web presence of software that guards people's money teaches you a professional level of paranoia. I recommend it.

  12. Engineering

    Three and a half years

    The company whose product is an oracle

    Chainlink builds the systems that tell blockchains the truth about the outside world - prices, events, facts. The industry's own name for them: oracles. I spent three and a half years building the sites and tools its developers rely on - three of those years while still running the web at ConsenSys, both engagements flowing through the research company. By then the joke was writing itself: my day job was helping machines report facts with sources, and so was everything else.

  13. Act III · The machines arrive
  14. Engineering

    2023

    A machine in the house

    Everything the career taught - connecting information, making it findable, running systems that do not lose things - pointed at one place, and it was not a data center. It was a machine on my own shelf, running models I could open up and inspect, holding data that never leaves the building. I stopped renting my computing and started owning it.

  15. The rule that runs through everything

    An answer arrives with the source it came from - or it does not arrive.

    I found the rule on the oldest shelf in the world, and I wrote it into the code.

  16. Engineering

    2026

    The year of the machines

    When the AI tools matured, two decades of collecting became fuel. In one year I shipped more new projects than in every earlier year combined: maps of knowledge, honest tests, tireless helpers, and an entire fleet of teaching sites. All of it built and run by one person, with a rack of machines at home. What every engineer dreamed of - one person building like a team - finally exists. It rewards exactly the people who spent years preparing for it.

  17. Research

    Now

    The room where the voices answer

    The strangest thing I have ever built: I taught a machine to read the oldest library I know, and to answer the honest way - name where you read it, or say nothing. Eight voices now answer from that library, out loud, from a machine in my house, and anyone can talk to them. It is my grounding testbed: the hardest citation problem I could find, running in public.

  18. The rule, measured

    Does the rule hold?

    It holds. Every answer must name a real source you can go and look up, verified word for word against the actual text - and if no source exists, the system says nothing. Only a model that clears every check ships.

    That discipline - cite or stay silent, measured, in public - is what I want to bring to a team working on AI you have to be able to trust. The research behind it is at jew.tech; the productized version is Digital Twin Pro.

What I build

Systems that have to show their work

One rule runs through everything: an answer arrives with the source it came from, or it does not arrive. Everything below runs today, in public, on hardware I own.

Maps of knowledge · honest AI

A library that answers with its sources

A living map of an ancient library - every piece carrying its source, every author joined to the teachers who taught them. Ask it anything, and the answer arrives with a reference you can look up. Then you can go check it.

Live now as OpenRabbi - a speaking agent grounded in the graph, running from my house.

Evaluation · testing

Every citation verified, word for word

Every citation is verified word for word against the real text before anything ships, in written tests and again in live conversation. You can read the methodology, run the queries behind every number, and check any figure yourself.

Infrastructure · product

A fleet, run like production

Forty-nine live sites built, deployed and QA'd by pipelines I wrote, on a GPU cluster in my house - and Digital Twin Pro, the company I founded: a private AI appliance you buy once and own outright.

Underneath all of it: machines in my own house. A system built to show you where its answers came from should be one you can open up and look inside. The personal research this grew out of is at /torah/.

Built and running

The sites

49 live sites. Every one of them started as something I wanted to understand well enough to explain out loud. Half my career happened in classrooms nobody called classrooms - teaching calculus, mentoring interns, building learning platforms. These sites are that same instinct with better tools, and every one of them is open to you. Here are six to start with; the rest, with what each one is for, live at jew.tech.

Connect

One conversation I always take

Are you building AI that has to be trusted before it is believed? Do you need someone who ships, measures, and publishes the receipts? Do you want an engineer who has run this exact discipline in production for years, not proposed it in a slide?

I am looking for my next role: AI engineering, grounded systems, the alignment problem in production. That is the conversation I always take. Write to me and I will answer - with sources!

Want the full picture first? The research is at jew.tech, the receipts at jew.tech/evidence, and the experiments I want to run next at jew.tech/agenda.