Executive summary. On Friday, July 24, 2026, between 1:00 and 3:00 p.m., the Coyoacán borough and Transgenia convened local entrepreneurs at Casa de las Letras in Mexico City. The workshop was not a keynote lecture: every attendee completed a 27-question digital-maturity self-diagnostic and joined a live demo of an AI tool that runs on Windows without a permanent cloud connection. Of the 50 seats we had mapped, 21 people arrived and 13 completed the diagnostic. The overall average was 27 out of 100, with the weakest dimension being Strategy (16/100). That is the real gap this piece documents.
1. The event: two hours, three blocks, one goal
The full name of the workshop was "AI and digital tools to scale your business without heavy investment." We designed it around a single operating principle: AI proposes; you approve. No promised dashboards and no distant futures — the goal was for each person to leave with three things in hand: a numerical snapshot of where their business stands today, a tool downloaded on their computer, and a first working prompt for the problem that actually hurts them.
We opened with the framing block, led by Efraín Carreón Ortiz, CEO of Transgenia, together with the person signing this piece. Danild Johan Zambrana, our technical lead, then took the second block — the hands-on part: how to install and light up a local AI agent on Windows in under thirty minutes. The team accompaniment was completed by Jaime Lozano Castañeda (CMO) and Julián Vega Herrera (Sr. Project Manager).
The event was made possible by the coordination with the Coyoacán borough and, in particular, by the support of Carlos Sanabria, whom we thank at the end of this text. The venue, Casa de las Letras, did its part: it is a space designed for conversation, not for showing slides in single file.
2. The flow: diagnostic first, tool second
Order matters. Before showing a single tool, we asked each attendee to complete a short self-diagnostic — five to seven minutes — with twenty-seven questions grouped into four dimensions: Customers, Strategy, Innovation, and Security. The result was not a vanity number; it was a per-dimension read plus a stage: Foundation (0–33), Traction (34–66), or Speed Up (67–100).
That diagnostic — which we simply call "knowing which rung of the ladder the business is on" — served two purposes. First, to lower the anxiety in the room. A lot of AI conversations start with "I'm not sure this is for me"; the diagnostic turns that question into an observable answer. Second, to steer the demo that followed: the prompts, the use case, and the sample tools were chosen to match the most frequent pain points reported in the first rows of the diagnostic.
3. The 13 diagnostics: what we found
Thirteen people completed the diagnostic. One of those responses is from our team (internal reference) and is not counted in the aggregates. With the 12 diagnostics from external entrepreneurs, the picture is clear.
| Stage | Entrepreneurs | Overall average (0–100) |
|---|---|---|
| Stage 1 · Foundation | 4 | 19 |
| Stage 2 · Traction | 4 | 18 |
| Stage 3 · Speed Up | 4 | 44 |
| Total | 12 | 27 |
Two observations worth saying out loud. First: even entrepreneurs who place themselves in Speed Up average 44/100 on the independent diagnostic. When someone has been scaling for a while and the number still doesn't cross the halfway line, the conversation stops being "I need more marketing" and starts being "I need more structure." Second: the gap between Foundation (19) and Traction (18) is essentially zero. That transition doesn't happen on its own; it requires a different kind of work, not more of the same.
Broken down by dimension, the gap gets sharper.
| Dimension | Average | Reading |
|---|---|---|
| Customers | 38 | The least weak dimension; there is market contact, even if scattered. |
| Innovation | 36 | Intent to try new things; execution is still intermittent. |
| Security | 18 | Data protection, formalization, and operational continuity: almost non-existent. |
| Strategy | 16 | The largest and quietest gap: without a written plan, there is no measurement. |
The headline is not "Mexican entrepreneurs don't know about technology." The headline is that Mexican entrepreneurs are moving faster on Customers than on Strategy. That asymmetry, inside a small business, is what keeps AI from paying off: a model that optimizes an operation without a written plan ends up automating the chaos.
4. Johan's demo: local AI on your desktop, without heavy investment
Danild Johan Zambrana anchored the second block on another operating principle: what runs on your computer is yours; what runs on someone else's cloud, less so. We showcased ZCODE — a desktop application from z.ai — configured to invoke both the z.ai GLM model and OpenAI's ChatGPT-5 nano, depending on what suits each task. Installation didn't require a top-end machine or an enterprise credit card; it ran on the same laptops the room had brought with them.
The live use case we built was generating a financial report for the past week from a simple spreadsheet: income, expenses, receivables. The AI read, summarized, proposed categories, and left the numbers ready for the entrepreneur to approve before saving. That was the point of the demo: not to replace the entrepreneur, but to remove the first hour of Monday-morning mechanical work. AI proposes; you approve. Same rule as with accountants, lawyers, or vendors: you delegate the draft, not the judgment.
Eleven people followed the installation step by step during the session. That is not an adoption metric — we are not pretending it is — but it is a signal that when the barrier is lowered to "download it and paste the prompt," the "I don't know about technology" defense loses ground quickly.
5. On video: moments from the workshop
Four clips recorded during the session, published on Transgenia's YouTube channel. The first is a walkthrough of the event; the next three are Shorts capturing specific moments of the dynamic.
6. What's next: mentoring, community, and the public demo
To close the technical block we walked through the ceiling: how far the free tier of a tool like ZCODE takes you, and what comes after. For businesses that have already crossed into Traction or Speed Up, the answer is the Transgenia Mentoring Program: a structured engagement to take AI from an individual desktop into a real business workflow, with Odoo as the backbone when the operation calls for it.
For those who want to see that next step before committing, we keep the Transgenia public demo open: a gated environment (access by verified corporate email) where you can play with the same kind of agent the session showed, already connected to simulated operational data. The demo shows a fixed "simulated data" badge on screen; none of the figures there belong to a real client.
7. Acknowledgments
This event existed because of two hands that do not appear on the poster. First, the Coyoacán borough, which opened Casa de las Letras and carried the invitation into the local entrepreneurial community; public spaces that work as knowledge plazas are a rare asset and should not be taken for granted. In particular, we thank Carlos Sanabria — whose institutional link is available in the Coyoacán borough directory — for holding the coordination together from well before the date, and for defending a format that favors conversation over exhibition.
Second, we thank the internal Transgenia team that made the operation possible: Efraín Carreón Ortiz (Direction), Jaime Lozano Castañeda (CMO), Danild Johan Zambrana (Technical Lead), Julián Vega Herrera (Sr. Project Manager). And, above all, the 21 people who showed up and the 13 who completed the diagnostic. An honest data point of 27 out of 100 is worth more, for decision-making, than an optimistic view without numbers.
If this day confirmed anything, it is this: the real gap is not between starting a business with AI and starting one without AI; it is between starting a business with a written strategy and starting one without it. AI accelerates what is already clear; the rest, we still have to write by hand.
Víctor Hugo Estrada Bello
Director of Projects, Transgenia
LinkedIn