Our verdict
Xmartlabs has built software since 2012, including CipherHealth’s hospital rounding app Orchid, and now focuses on AI products for healthcare, from clinical workflow tools to medical models that run on the device (homepage; CipherHealth case). It suits a US healthtech company that wants a senior team in a nearby time zone to design and build an AI product under HIPAA rules, and it is less suited to a buyer that wants back-office automation or data platform work outside healthcare.
The right buyer
- Healthtech startups and scale-ups. It has been CipherHealth’s long-term product partner, building Orchid, its best-selling product, and the Echo suite for hospitals with millions of patients (CipherHealth case).
- Products where patient data cannot leave the device. For Fasten Health it benchmarked medical language models under 8 billion parameters and designed a retrieval system that can run locally or in a hybrid setup (Fasten Health case).
- Teams that need a quick first version. Its AI Studio Week is a five-day sprint that takes a healthtech idea to a working product (services).
Reasons to look elsewhere
- Its focus is healthcare, so buyers in other industries will find fewer recent cases.
- Its clients are mostly US product companies, so European buyers will find fewer local references.
Xmartlabs, At A High Level
Xmartlabs was set up in Montevideo in 2012 and works from Bulevar España 2529, and it also has an office in San Francisco (Uruguay XXI; about page). Agustín Hernández Bonaglia, its co-founder, is chief executive (Uruguay XXI). Its open source iOS libraries, such as Eureka and XLPagerTabStrip, have been built into more than 40,000 apps, and it published Bender, a framework for running neural networks on iPhones, before Apple’s own tools ran models on dedicated chips (about page).
- Headquartered in Montevideo, Uruguay, with an office in San Francisco (about page).
- More than 100 engineers across five locations by its own count; we estimate annual revenue at $6 million to $8 million (homepage).
- Services cover a healthcare AI product studio, machine learning and data engineering for healthtech, on-device and edge AI, and senior pods embedded in client teams (services).
- Partner badges on its site include AWS and Anthropic’s Claude partner network (homepage).
Scores by criterion
We scored Xmartlabs from its homepage, about, services, and case study pages and its listing with Uruguay XXI, the national investment agency. Production track record scores highest; commercial clarity scores lowest because it publishes no terms.
| Criterion | Weight | Score | Basis |
|---|---|---|---|
| Production track record | 20% | 7.5 | 14 years of shipped products; long-running CipherHealth work and AI cases for Fasten Health and Omniscia. |
| Technical depth | 15% | 7.0 | On-device AI, medical model evaluation, RAG, and mobile engineering with widely used open source. |
| Industry depth | 10% | 6.5 | Deep healthcare focus, plus media, legal, and consumer apps. |
| Delivery model and handoff | 15% | 7.0 | From discovery and design through build, hardening, and handoff. |
| Commercial clarity | 10% | 5.0 | No published rates or minimums. |
| Platform independence | 10% | 6.5 | Partners with AWS and Anthropic; local and hybrid model options. |
| Stability and scale | 10% | 6.5 | Founded 2012; more than 100 engineers. |
| Independent evidence | 10% | 6.5 | Attributed client quotes and open source projects used in more than 40,000 apps. |
| Overall | 100% | 6.7 | Weighted average |
Where Xmartlabs fits
Xmartlabs is used by US healthtech startups and enterprises that want an AI-native product designed and built by a senior Uruguayan team working in US hours, under HIPAA and clinical rules. On price it sits in the upper middle of Latin American product studios, below US agencies. Competitors on this site include Arionkoder and NaNLABS for Latin American AI product teams, and Rootstrap for nearshore product engineering.
| Need | Fit | Why |
|---|---|---|
| An AI product for patients or care teams | Strong | It builds care coordination, patient intake, and hospital rounding products. |
| Medical AI that runs on the device | Strong | It benchmarks small medical models and built Bender for running neural networks on iOS. |
| A first version in a week | Good | Its AI Studio Week is a five-day sprint. |
| Back-office automation outside healthcare | Weak | Its recent work centers on healthtech products. |
What clients and leaders say
Xmartlabs’s leaders and clients describe its work in published announcements. One buyer concern is written from public facts about the firm.
“Hands-down one of the best dev teams I’ve worked with. For building a business you need to balance cost with quality. Xmartlabs hits that equation out of the park.”
“What’s most impressive and unique about Xmartlabs is their ability to act as a true partner with aligned interests and not just one party to a transitional relationship.”
“Their expertise, particularly at the early stages of research and proposing possible solutions, was quite detailed and balanced. They look for the best overall solution.”
“The team understood HIPAA and built a strong product, but nearly all their recent cases are healthcare, so we checked their fit for our non-clinical features before signing.”
Where customer reviews couldn’t be found, our analysts attempted to give a voice to buyer sentiment, drawing on real information from review sites, forums and product documentation. Based on: the focus areas on Xmartlabs’ homepage and its case studies.
Who leads Xmartlabs
Agustín Hernández Bonaglia co-founded Xmartlabs and is its chief executive (Uruguay XXI).
