ai-training
(19)
Órbi began with a bank account. Now it’s betting on a whole neighborhood.
A startup community, a bank, a builder and a car-rental company gave Belo Horizonte a place to meet. Órbi’s next act brings research, executive education and the streets of Lagoinha into the same conversation.
PALO IT wants to fix the work around the code
The technology consultancy is putting AI to work on the handoffs, habits and legacy systems that slow software down. Its projects with airlines, health insurers and fitness clubs show what that bet looks like in practice.
Infinite Lambda Wants to Fix AI’s Plumbing
Before an AI agent can give a useful answer, somebody has to sort out the data. Infinite Lambda turns that unglamorous job into software, migration projects, and teams that can run the result.
QA Apprenticeships puts the day job on the syllabus
A Costa Coffee apprentice forecasting cup waste captures QA’s pitch: learn a technical skill, then put it to work. The harder part is making room for learning inside a working week.
Beround Spent a Decade Hunting Software Bugs. Now It Wants to Fix AI Adoption Before the Tools Arrive
The Dutch consultancy built its reputation on the unglamorous discipline of testing. That same habit now powers a people-first AI playbook that starts small, measures what matters and leaves room for a human to say no.
MangoBoost Wants to Stop Your Million-Dollar GPUs From Waiting
The four-year-old semiconductor startup began with a card that takes chores away from CPUs. Now it is selling the whole kitchen - silicon, storage, servers and software - to make AI infrastructure cheaper to run and harder to strand.
The Startup Teaching Horses to Talk Back
A horse can't tell you its left hind is off. Equestic's SaddleClip listens to the gait instead - and riders in 65 countries are now training to what the data says.
The AI Consultancy That Wants to Make It Past the Pilot
Most companies do not need another AI demo. Artefact has built a 2,500-person consulting business around the less glamorous job of turning scattered data, stubborn workflows and wary employees into systems that get used.
CoreWeave
The crypto-mining startup that became the picks-and-shovels supplier of the AI boom - renting Nvidia GPUs to the labs building the future, one megacluster at a time.
Trainual
An all-in-one employee training platform that connects knowledge, systems, roles, and processes to deliver instant answers, job-ready training, and confident execution for every employee.
CueBench
CueBench builds reinforcement-learning environments for frontier AI labs, focused on scientific reasoning and performance engineering - the kinds of hard, specialized tasks where the best models still fail more than 80% of the time. The company's bet is that the next gains in model capability will come less from raw pre-training and more from training agents inside carefully constructed environments that force them to learn the things they are currently bad at. Founded in 2026 by Dillon Mehta, Rishan Hemrajani, and Neel Gadde, CueBench is part of Y Combinator's Summer 2026 batch and is backed by NVIDIA Inception.
TensorPool
TensorPool (YC W25) is a San Francisco startup that turns cloud GPUs into a git-style command line. Instead of wrestling with SSH sessions, provider dashboards, and idle-instance billing, machine learning engineers describe a training job in a config file and run 'tp job push' - TensorPool spins up a multi-node cluster across partner clouds, streams logs, saves outputs, and shuts everything down, billing only for runtime. It advertises GPU access at roughly half the cost of major cloud providers, with H100s starting around $1.99/hr. Founded by three Stanford dormmates, the company was accepted into Y Combinator's Winter 2025 batch.
Deeptune
Deeptune is a New York AI company that builds high-fidelity simulation environments - 'training gyms' - where AI agents practice real digital work such as accounting, customer support, DevOps, and spreadsheet and CRM tasks before being deployed. Its reinforcement-learning environments recreate the day-to-day software of roles like accountants, lawyers, and engineers so agents can learn multi-step 'computer use' the way a pilot trains in a flight simulator. Founded in 2025 by Tim Lupo and Lukas Schmit, Deeptune raised a $43M Series A led by Andreessen Horowitz in March 2026 and was acquired by AI unicorn Mercor in July 2026.
Section
Section is a New York-based AI workforce-transformation company that helps enterprises turn expensive AI licenses into measurable behavior change. Founded in 2019 by NYU Stern professor Scott Galloway as the business-education startup Section4 and led by CEO Greg Shove, the company pivoted hard into AI in 2023. Its software - Section HQ, Section Coach and the ProfAI assistant - pairs certification, coaching and use-case discovery to move employees from 'I have a ChatGPT login' to actually applying AI on the job. Customers include Johnson & Johnson, Nike, Unilever, Okta and Horizon Media.
100mentors
100mentors is an Athens-based company that helps organizations adopt generative AI - with people, not just tools, at the center. Founded in 2015 as an on-demand mentorship platform that connected 21,000+ students across 17 countries with 2,500+ mentors from top universities and companies, it pivoted with the rise of generative AI into an enterprise partner offering AI upskilling bootcamps (GenAI Academy), an AI-powered meeting tool (Wiserwork), and the GenAI Summit, positioned as one of Europe's largest generative-AI leadership conferences.
Starcloud
Starcloud is a US deep-tech company building data centers in orbit. By placing GPU clusters in space, it taps near-continuous solar power and passive radiative cooling to run AI training and inference without the land, water, grid-permitting, and energy constraints that bottleneck data centers on Earth. In November 2025 its Starcloud-1 satellite became the first spacecraft to train a large language model in orbit using an Nvidia H100 GPU; its $170M Series A in March 2026 made it the fastest unicorn in Y Combinator history.
Zeromatter
Zeromatter builds the simulation infrastructure that autonomous-systems companies use to develop, test and validate their products before they ever touch the real world. Founded by former Tesla Autopilot simulation lead Ian Glow, its platform combines high-fidelity sensor simulation, automatic environment generation, multi-agent co-simulation frameworks, execution infrastructure and developer tooling into one system - aimed at making simulation a default tool for autonomy, aerospace, automotive, agriculture, drones and green energy.
Manara
Manara is a Silicon Valley-backed edtech company training and placing software engineers, AI and cloud talent across the Middle East and North Africa. Through cohort-based learning, mentorship from senior engineers at companies like Google and Meta, and partnerships with AWS, Manara has trained over 300,000 learners and helped hundreds land jobs at global tech firms.
来Offer (LaiOffer)
LaiOffer is a Silicon Valley-based career and technical training company founded in 2013 that prepares software engineers, data scientists, and AI practitioners for jobs at top tech companies. Through project-based courses, mock interviews, and mentorship from FAANG engineers, it has helped more than 8,000 candidates land roles at Google, Meta, Amazon, Microsoft and similar firms.