ylemlab. Book a call
Part
Edgar Rincon
Role
Forward Deployed Engineer
Issued by
YLEM Lab
Status
Taking on projects

I embed with your team and build production software on your floor.

Manufacturing and AI. I scope the problem, build the fix inside your environment, deploy it, and hand it over running.

Book a call See my work: OnFaktor and Stud Labs

What I build for you

  1. 1Camera inspection and countingCameras that count and judge every part, and keep the picture as proof.
  2. 2Downtime and quality trackingEvery stop and every reject recorded on its own, with the reason.
  3. 3AI on the factory floorAssistants that read your procedures and answer on the line. In the cloud or on your own server.
  4. 4ERP and Odoo integrationProduction numbers that reach the ERP without anyone retyping them, so the cost shows up where finance looks.
  5. 5Robots and automationCameras, robots and shop-floor systems made to work with each other.
  6. 6Data productsA scattered body of data turned into a tool people use. Stud Labs does this for horse racing.
Technical drawing of the Ylem atom: a core, three satellites and a base. Six numbered callouts point to the core, the lower left satellite, the copper satellite on top, the lower right satellite, the bonds and the base. YL-001 1 2 3 4 5 6
Part YL-001: the Ylem atom. Tap or hover a service to find its callout.

How the work travels

A route card goes with every part through a plant. This is mine for a project: five operations, each signed off before the next.

OP 10 Embed

  1. OP 10

    Embed

    I join your team on site. I watch the line, talk to the operators and read the reports they do not trust.

    Signed
  2. OP 20

    Scope

    One problem, written down in plain words, with the finish line agreed before I build anything.

    Signed
  3. OP 30

    Build

    Inside your environment, on your data, with your security rules. No prototype that only works on my laptop.

    Signed
  4. OP 40

    Deploy

    On your floor, running, with the first real shift watched together.

    Signed
  5. OP 50

    Hand over

    Your people run it. The documentation is written for the next shift, not for me.

    Signed

Work I have built

Each record has the same four lines: the problem, what I built, what it does now, and where to see it.

The OnFaktor Station live screen: a camera image of a moulded part marked OK, with the cycle time, downtime and shift counts beside it
OnFaktor Station, live screen. A demo plant with synthetic data.

OnFaktor

One box that arrives with the whole suite running.

Problem
Plants run on reports, not on what happened. A line stops and nobody records how long, or why. Parts come back and nobody counts what they cost.
Built
The OnFaktor Station: a camera counts every part and every stop, the operator says why, Kaizen turns the losses into a better procedure, and the ERP shows the money. OnFaktor AI answers on the line.
Now
Running, with a live demo. The AI runs in the cloud, on a GPU hosted for the customer, or on the customer’s own server.

Stud Labs

Data tools for pinhooking. Four apps that follow a yearling from the sale where it is bought to the sale where it is resold.

Problem
Much of the horse trade runs on intuition, and the data often contradicts it.
Built
What to buy, what it should resell for, what each horse costs, and how it is raised, each in its own app, one after the other.
Now
Four working apps, under an exclusive contract with Rondon Stables. Stud ADN stays on my own machine by design, because the data is the asset.

The screens show demo data.

  1. Stud ADN: the yearlings from one sale, each with its buy price and the profit it made on resale
    Stud ADN Is this yearling a good buy? Judged from what its sire’s and dam’s foals actually sold for.
  2. Stud Auctions: the picks shown to investors on sale day
    Stud Auctions The picks, shown to investors on sale day and checked against the real hammer price.
  3. Stud Control: the cost of each horse from purchase to sale, on a phone
    Stud Control What each horse costs, from purchase to sale, with receipts. Works offline on a phone.
  4. Stud Dynamics: a daily log of how each yearling is raised, next to its breeze times
    Stud Dynamics How each yearling is raised, day by day, set against its breeze times.

YLEM Lab

The studio behind this work. It is where customer projects are taken on. One person who builds, so the person you talk to is the person who writes the code.

From the courtroom to the factory floor

Most engineers cannot copy this part of the story. I learned to read a contract before I learned to read a line.

Revision history
RevWhat changed
ATrained and worked as a lawyer. Learned to find the sentence that decides everything.
BBecame an engineer. Software first, then cameras, robots and AI.
CTook the work onto the factory floor, where a bug stops a line and not a screen.
DBuilt OnFaktor and Stud Labs end to end, and founded YLEM Lab.

ylem.

noun

  1. The first matter. The stuff everything is made from, before it has a shape.
  2. My family. The first letters of four people, in a row. Every project I build starts with them.
Four particles. One family.

Where the word comes from

  1. Ancient Greece

    Hylē

    Aristotle writes about hylē, the matter that things are made of.

  2. The Middle Ages

    Ylem

    Writers in Latin and English use ylem for the first matter, before anything has a shape.

  3. 1948

    The first moments

    Physicists Ralph Alpher, Robert Herman and George Gamow use ylem for the hot, dense matter at the start of the universe.

  4. Today

    Ylem Lab

    A place where things start. Software built from the first line to the factory floor.

What people ask first

What is a Forward Deployed Engineer?

An engineer who works inside the customer’s own environment instead of from a distance. I scope the problem, build the software where it will run, deploy it, and stay until it works.

Do you only work in manufacturing?

Manufacturing and AI are where I work most. Stud Labs shows the same method on a different kind of data. If your problem is a scattered pile of data, ask.

How does a project start?

With a call. You tell me the problem. If I can solve it, I say how I would go about it. If I cannot, I say that too.

Is OnFaktor part of working with you?

OnFaktor is a product with its own page. A project can use it, extend it, or have nothing to do with it. Your problem decides.

Tell me the problem.

I will tell you if I can solve it, and how I would start.

Book a call

Released by Edgar Rincon