AI Week LATAM 2026: Mexico and Chile Join Colombia in Nvidia’s Regional AI Training Push

AI Week LATAM 2026 runs September 29 through October 3, 2026, and for the first time it is not a Colombia-only event. SoftServe and Nvidia have opened the third edition to Mexico and Chile as well, targeting more than 4,000 participants across the three countries, with 400 seats reserved for Nvidia Deep Learning Institute (DLI) certification.

Two prior editions stayed inside Colombia and trained roughly 2,200 people combined. This year’s jump to three countries, four university partners and a certification track that is deliberately smaller than the headline attendance number is the real story: it shows how a regional vendor-plus-university training model scales, and what it does and does not certify.

This piece covers the expansion mechanics, the university-partnership model behind it, why the certification funnel looks the way it does, and what any L&D leader watching Latin America’s AI upskilling market should take from it.

What is AI Week LATAM 2026, and when does it run?

AI Week LATAM 2026 is a free, week-long AI training program run by SoftServe and Nvidia with regional university partners, combining public talks with hands-on Nvidia DLI workshops. It runs September 29 to October 3, 2026, across Colombia, Mexico and Chile, with sessions delivered online, in person and hybrid.

The program is now in its third edition. The first two editions ran in Colombia only, according to SoftServe’s official announcement. Chile’s on-site kickoff at Universidad Técnica Federico Santa María in Valparaíso began a day earlier, on September 28, with in-person workshops on agent tooling and Nvidia Omniverse, per regional coverage from Diario Financiero’s report on the Chile program. The bulk of the pan-regional schedule, including the Nvidia DLI workshop block, falls inside the September 29 to October 3 window.

Why did AI Week LATAM expand from Colombia into Mexico and Chile this year?

The expansion happened because the Colombia-only model had already proven it could deliver: two editions and roughly 2,200 trained participants gave SoftServe and Nvidia a template worth repeating, and university partners in Mexico and Chile agreed to host their own local instance of it. Scaling a training program regionally usually means replicating a working delivery model, not reinventing one per country.

That is what happened here. Universidad de los Andes and Universidad EIA, the two Colombian hosts from earlier editions, kept their roles in Bogotá and Medellín. Tecnológico de Monterrey joined as the Mexico host in Guadalajara, and Universidad Técnica Federico Santa María joined as the Chile host in Valparaíso, according to Mexico Business News’s coverage of the 2026 expansion.

Adding two countries without adding new content tracks or a new certification structure is a specific operating choice. It suggests SoftServe treated the Colombia edition as a reusable playbook: same talks-plus-workshop format, same Nvidia DLI courses, same open-registration model, just handed to two new university hosts and their local student and engineer populations.

Which universities and cities are hosting AI Week LATAM 2026?

Four universities across three countries are hosting AI Week LATAM 2026 in 2026: two returning Colombian hosts and two new hosts in Mexico and Chile. The table below lists each host, its city and its role in the program.

Every host city also carries hybrid and fully online sessions for participants who cannot attend in person, and SoftServe describes the online talks track as open to other Latin American countries beyond the four host cities.

How does the university-partnership model differ from a vendor-run AI bootcamp?

A university-partnership model like AI Week LATAM ties training to an academic institution’s own student and faculty network and its credibility with local employers, where a pure-vendor bootcamp relies on the vendor’s own brand and marketing reach to fill seats. The practical difference shows up in who walks in the door and how the content gets validated.

A vendor-run bootcamp typically sells directly to individuals or to employers, sets its own curriculum, and issues its own completion certificate. That is fast to launch and easy to run identically in any country, but it depends entirely on the vendor’s outreach and reputation to attract participants.

AI Week LATAM instead routes registration and, for the DLI track, eligibility screening through the host universities, which supply the “active institutional academic email” requirement flagged in the source coverage. That gives the program a built-in participant pipeline (students and faculty already inside four large universities) and a layer of academic legitimacy that a standalone vendor bootcamp has to build separately. The trade-off is that the workshop track is gated to people with institutional access, while the open talks track stays available to anyone.

For a training platform buyer evaluating vendor partnerships in a similar way, the same logic applies to internal LMS selection: a platform’s AI-driven content and assessment tools are only as good as the delivery and credentialing model wrapped around them, and a partnership with an established academic or professional body often does more for adoption than the AI feature list alone.

What do the 400 Nvidia DLI certification slots actually get you?

The 400 Nvidia Deep Learning Institute slots at AI Week LATAM 2026 lead to an official Nvidia certificate tied to a specific technical course, not a general attendance certificate. Nvidia’s DLI issues these for completing hands-on workshops such as Fundamentals of Deep Learning, Rapid Application Development with LLMs, Generative AI with Diffusion Models, and Applications of AI for Anomaly Detection, according to the workshop list published on the official AI Week LATAM 2026 event page.

These are competency-based credentials. Nvidia’s own certification materials describe them as validating that a participant can apply a specific technical skill, not that they attended a session. That distinction matters for anyone using this event as an external signal of workforce readiness: a DLI certificate on a resume means the holder completed a defined, hands-on curriculum with prerequisites, not that they sat through a keynote.

Workshop eligibility is narrower than the talks track. Participants need an active institutional academic email address and, depending on the course, working knowledge of Python, linear algebra, probability and statistics, and a framework such as PyTorch or Keras. Nvidia and SoftServe review applications and confirm seats by email based on eligibility and available capacity, per the sourced coverage.

Why 400 certified seats against 4,000-plus participants? The funnel, explained

The gap between 4,000-plus total participants and 400 DLI certification slots is not a shortfall, it is the program’s intended structure: a wide, low-barrier awareness layer feeding a narrow, prerequisite-gated certification core. Roughly one in ten attendees is positioned to leave with a formal Nvidia credential; the other nine get exposure, talks and hands-on demos without a credentialing requirement.

What a roughly 10 percent certification rate tells you

That ratio is a reasonable design for a regional awareness campaign, and it is also a limit worth naming plainly. Mass reach builds a pipeline of people who know the vocabulary and have touched the tools; a much smaller group actually walks away with a portable, verifiable credential. Anyone reporting “4,000 people trained” without separating the two numbers is overstating what the majority of attendees actually completed.

This two-layer structure is common in corporate and workforce AI training more broadly, and it is one reason corporate training budgets for 2026 increasingly separate spend on broad awareness content from spend on certified, assessed learning paths. Treating the two as the same line item understates what certification actually costs to deliver at scale.

What does this expansion signal about Nvidia’s workforce strategy in Latin America?

Nvidia’s expansion of AI Week LATAM into Mexico and Chile signals a deliberate, university-anchored approach to building AI talent pipelines in Latin America rather than a one-off marketing event. Repeating a proven Colombia model in two new countries, through the same academic-partnership structure, points to a multi-year regional workforce strategy rather than a single campaign.

John Howard, SoftServe’s regional Latam lead, framed the goal in workforce terms rather than brand terms, saying the opportunity “lies in developing necessary talent and capabilities to deploy these technologies across industries where productivity, safety and efficiency are critical,” according to SoftServe’s press materials. That framing lines up with the program’s topic list: physical AI, agentic AI, digital twins, robotics and anomaly detection are all industrial and operational applications, not consumer-facing AI use cases.

For training platform vendors and buyers watching the region, this is also a distribution signal. A vendor that can get four universities in three countries to co-host a technical training week has solved a go-to-market problem that many corporate AI training providers still struggle with outside their home market.

Is AI Week LATAM 2026 free, and who can attend?

Yes, AI Week LATAM 2026 is free for all participants, and the talks and sessions track is open to students, engineers and anyone interested in AI, regardless of prior experience. The Nvidia DLI workshop track carries no fee either, but it has narrower eligibility than the open talks track.

Talks and expert panels are available online, in person, or in hybrid format depending on location, with no technical prerequisite. On-site sessions in Bogotá, Medellín, Guadalajara and Valparaíso have limited physical capacity, so early registration matters more for in-person seats than for the online track.

How do you register, and what do the Nvidia workshops require?

Registration for AI Week LATAM 2026 is a single online form covering both the open talks track and the Nvidia DLI workshop applications; workshop applicants additionally need an active institutional academic email and must meet each course’s stated technical prerequisites. Confirmation for workshop seats is sent by email based on eligibility and remaining capacity, not on a first-come basis alone.

Because workshop demand is expected to exceed the 400 available slots, applicants without an institutional email, or without the listed prerequisites for a given course, should expect to be routed toward the open talks track instead. Nvidia’s own DLI course catalog lists the technical bar per course, which is worth checking before applying for a specific workshop rather than after being declined.

What topics does the 2026 curriculum actually cover?

The 2026 curriculum spans physical AI, agentic AI and generative AI, plus applied tracks in CUDA, machine learning, autonomous agent deployment, digital twins and robotics, and anomaly detection. Chile’s on-site program adds sessions on coding agents, agent security, conversational AI and Nvidia Omniverse, based on the schedule reported by Diario Financiero.

This is a deliberately industrial slant rather than a general “intro to AI” curriculum. Topics like anomaly detection and digital twins for robotics are aimed at engineers who will apply AI inside manufacturing, logistics or industrial operations, not at a general office-worker audience learning to use a chatbot.

What should L&D teams outside Latin America take from this?

For L&D leaders outside the three host countries, AI Week LATAM 2026 is worth tracking as a live case study in how a vendor-university training model scales regionally, and as a reference point for what a credible AI upskilling funnel looks like: broad free access at the top, a certified and prerequisite-gated track underneath it.

Teams evaluating their own AI training stack can use the same funnel logic internally: publish accessible, low-barrier AI literacy content broadly through your learning platform, then reserve certification or assessed learning paths for the subset of staff who actually need a verifiable, role-specific AI skill. Treating every AI training initiative as if it should produce a certified outcome for 100 percent of participants is neither realistic nor necessary, and this program’s own numbers make that point without anyone having to argue it.

Conclusion

AI Week LATAM 2026 turned a two-edition, Colombia-only program into a three-country regional initiative by repeating a working delivery model rather than redesigning it, and its 400-seat certification track against 4,000-plus participants shows a funnel structure worth recognizing in any large-scale AI upskilling effort.

If you run L&D or training operations and want to track how this scales into 2027, or how the certification completion numbers compare to the registration figures once the week wraps on October 3, that follow-up data point will tell you more about program quality than the initial headline count did.

Start by checking whether your own organization’s AI training already separates broad awareness content from prerequisite-gated, certified learning paths, since that structure is exactly what made this program’s numbers make sense.