Top 10 AI Startups to Watch in Colombia in 2026

Table of Contents

• Colombia's AI tasting flight
• Arkangel AI
• Vozy
• HoyTrabajas
• Treinta
• CORAL Robotics
• WSeeds
• SOLYON Technologies
• WeKall
• Loto Latam
• NanoFreeze
• Final Sip: Training Your AI Palate
• Frequently Asked Questions

Arkangel AI

Arkangel AI sits where Colombia’s health gaps meet its fast-growing data science talent. Co-founded in Bogotá by Laura Velásquez and Jose Zea, the company turns messy clinical records and claims data into diagnostic models that hospitals and insurers can actually deploy. Listed among the top local AI ventures on the F6S directory of Colombian AI companies, Arkangel has raised a Seed round that includes international investor 468 Capital, positioning it as one of the country’s most visible healthtech bets.

Why it matters

Colombia’s mixed public-private health system generates mountains of underused data, while specialist access remains uneven outside Bogotá and Medellín. Arkangel offers an AI-as-a-Service platform that lets providers build and validate disease-specific models without writing code, a critical advantage for IPS and EPS that lack in-house ML teams. According to its own metrics and investor materials, Arkangel’s models can be up to 15x more cost-effective and 10x faster to deploy than traditional approaches. The platform is already used by pharmaceutical companies, insurers, and clinics for early disease detection, and has been highlighted in independent rankings such as DesignRush’s overview of AI companies in Colombia.

• No-code pipelines for building and updating diagnostic models
• Tooling for data cleaning, validation, and bias checks in clinical datasets
• Deployment options that fit both on-prem and cloud-constrained hospitals

What to watch in 2026-2028

The next phase is less about prototypes and more about navigating regulation and scale. As Colombia refines rules around health data and algorithmic accountability, Arkangel is well-placed to become a reference partner for pilots under initiatives like ColombIA Inteligente 2026 and for regional health insurers looking for standardized AI tooling.

• Regulatory momentum in data protection and medical-device-grade AI
• Regional expansion into Peru, Ecuador, and Central America’s similar health systems
• Exit potential via MedTech, big cloud, or global pharma seeking Spanish-language clinical AI expertise

Vozy

Anyone who has shouted “asesor humano” into a crackly IVR knows how broken corporate phone menus can be. From its base in Medellín’s growing SaaS corridor, Vozy is rebuilding that experience with “Lili,” a conversational AI tuned to understand eight different Spanish accents across Latin America.

Why it matters

Customer support is one of the biggest cost centers for Colombia’s banks, telcos, and utilities, and it is also one of the easiest places to see AI’s impact on the P&L. According to regional startup analyses, Vozy’s assistant can resolve around 50% of customer service issues on the first attempt, and large enterprise clients report 40%+ efficiency gains once Lili is fully integrated into their contact centers. Unlike simple FAQ bots, Vozy connects deeply into back-office systems to handle flows like payment promises, plan changes, or appointment scheduling, and it already manages thousands of concurrent calls for major LatAm enterprises highlighted in Contxto’s list of regional AI startups to watch.

• Speech recognition optimized for multiple Spanish accents, including Colombian, Mexican, Caribbean, and Southern Cone
• End-to-end automation of calls, WhatsApp, and web chat in a single orchestration layer
• Analytics on intents, drop-off points, and satisfaction to continuously retrain models

What to watch in 2026-2028

Medellín’s dense BPO and nearshore ecosystem makes Vozy strategically important. As more contact centers serving the U.S. and LatAm consolidate around the city, Lili can become the default AI layer for hybrid teams where some agents sit in Sabaneta and others in Mexico or Peru.

Key questions for the next two years include how deeply Vozy can embed into CRM and ticketing stacks, whether it can secure multi-country deals with BPO giants, and if its growth trajectory pushes it toward a late-stage funding round or makes it an attractive acquisition for global CX or cloud communications platforms looking for a native Spanish-language AI asset.

HoyTrabajas

In Bogotá’s retail and logistics corridors, hiring for high-turnover roles can feel like an endless avalanche of WhatsApp messages and walk-in CVs. HoyTrabajas was built for that reality. The company, founded in Bogotá, uses AI agents to pre-screen and vet thousands of candidates and deliver shortlists to employers in under 24 hours.

Why it matters

Operational jobs in commerce, logistics, and call centers still drive much of Colombia’s urban employment, but HR teams often sift manually through hundreds of low-signal applications. HoyTrabajas attacks this bottleneck with a stack that goes beyond keyword matching. Its models:

• Parse CVs, chat replies, and audio responses from non-technical candidates
• Score fit against job requirements in high-turnover industries
• Automate communication, reminders, and interview scheduling at scale

According to revenue data compiled by GetLatka’s profile of HoyTrabajas, the company reached about US$7.4M (~29.7B COP) in revenue with a 48-person team, a strong signal that the automation thesis is resonating with employers across Colombia and the wider region.

Traction and capital

Early on, HoyTrabajas raised roughly US$430,000 (~1.6B COP) in Seed funding, capital that has gone into product refinement and regional expansion rather than an oversized payroll. That lean profile matters in a market where HR tech buyers scrutinize ROI line by line. The platform’s promise - pre-screened operational talent in under a day - aligns well with nearshore call centers and logistics operators scaling out of Bogotá, Medellín, and the Caribbean ports.

What to watch in 2026-2028

• Deep integrations with large retailers, 3PLs, and BPOs that run thousands of simultaneous searches
• Regulatory debates around algorithmic bias, data protection, and transparency in hiring decisions
• Strategic buyer interest from global HR suites and regional employment platforms seeking a proven AI-native pipeline for frontline roles

Treinta

Across barrios from Soacha to Bello, Treinta began as the simplest possible promise: a free digital ledger so micro-businesses could track who owed what. Today, the company operates between Bogotá and Medellín and serves millions of small merchants across the region, making it one of the most widely adopted SMB tools in Colombia. Analysts at StartupBlink’s ranking of Colombian startups consistently list Treinta among the country’s standout growth stories.

From notebook to Idealane

In 2025, Treinta moved beyond bookkeeping with the launch of Idealane, an AI-driven builder that lets small businesses create apps and websites via chat. For a tienda de barrio in Ciudad Bolívar or a peluquería in Itagüí, that means skipping agencies and simply describing what they need - “a catalog with prices, WhatsApp button, and delivery zones” - and letting the AI assemble the pieces.

• Auto-generates simple apps, catalogs, and landing pages from natural-language prompts
• Connects to sales and inventory data already flowing through Treinta
• Suggests promotions or bundles based on observed purchase behavior

Why it matters in Colombia’s SMB-heavy economy

Fintech has become the largest single startup vertical in Colombia, but many micro-merchants are still stuck between being banked and being truly digital. Treinta sits in that gap. By combining cash-flow tracking with a conversational AI builder, it gives owners a practical way to move online without hiring a developer or learning complex tools.

Strategically, that “fintech + AI tools” bundle is powerful. Whoever owns the daily interface with tiendas can later layer on digital payments, working-capital loans, insurance, and even last-mile logistics. For engineers and product builders in Bogotá and Medellín, Treinta’s arc shows how an unglamorous ledger app can become a platform - once you add an AI layer that listens to what small businesses actually want to build.

CORAL Robotics

Just north of Bogotá, in the highland municipality of Cajicá, CORAL Robotics is rethinking how Colombia’s flagship crop is graded. Instead of relying only on human pickers and visual checks, CORAL builds robots that use spectral sensing to analyze the internal quality of each coffee cherry in real time, moving beyond standard RGB cameras.

This matters for producers in Huila, Antioquia, Nariño, and beyond, where small and medium growers often lack access to advanced post-harvest tech. CORAL’s hardware-software stack gives them a way to compete with industrial plants without abandoning origin stories or quality premiums.

• High-precision classification by reading spectral signatures correlated with ripeness and defects
• Automated sorting that boosts consistency and usable yield per lot
• Historical dashboards tracking quality over harvests and climate conditions

Founded by William, Jhojan, and Nicolás in the greater Bogotá region, CORAL has attracted angel backing from investors such as Adonis Tupac Ramírez Cuéllar. Its progress from lab prototype to farm-ready hardware is one reason databases like the Cuantico VP report on Colombian startups flag coffee- and climate-focused ventures as a distinct national strength.

The bigger play is climate-smart agriculture. Once spectral and robotics infrastructure is in place for coffee, the same platform can extend to crops like cacao and avocado that face similar post-harvest bottlenecks. As Latin American investors profile Colombia as a launchpad for the region’s food-tech and agritech wave, highlighted in analyses from outlets such as LatamList’s coverage of outstanding Colombian startups, CORAL sits at the intersection of data-rich agriculture and export-ready innovation.

For ML engineers and roboticists in Bogotá and Medellín, CORAL is a reminder that some of the most interesting AI problems in Colombia are not in app stores, but in the drying patios, beneficio plants, and cooperatives that still define the rural economy.

WSeeds

On many Colombian farms, the most advanced “system” is still a worn notebook in a plastic bag. WSeeds was created for that reality. From Bogotá, the team built Agrochat, a WhatsApp-native AI assistant that lets smallholders log activities with simple text or voice notes and quietly turns that stream into structured agronomic data.

WhatsApp as the operating system of the campo

Rather than asking farmers in Huila, Cauca, or Magdalena to learn new apps, WSeeds meets them where they already are: WhatsApp. Agrochat is designed for low-connectivity environments and voice-first users, translating recordings into organized records of planting dates, input usage, labor costs, and harvests.

• Voice-first capture for farmers more comfortable talking than typing
• Offline-friendly workflows that sync when a signal appears
• Automatic dashboards showing yields, costs, and plot performance over time

From field notes to financial identity

Analysts at Cuantico VP highlight WSeeds as a key piece in end-to-end farm digitization, noting its potential to become a de facto “data layer” for small producers across the Andean belt. Backing from institutional actors like the Cardano Foundation signals that global blockchain and impact investors see smallholder data as an asset class, not a byproduct. In parallel, ecosystem overviews such as Productos AI’s analysis of Colombia as a nearshore AI hub point to agtech pilots like WSeeds as proof that Colombia can incubate AI products for emerging markets far beyond its borders.

What to watch next

The real leverage comes when Agrochat data is accepted beyond the farm gate. Over the next few years, WSeeds’ growth will hinge on partnerships with banks, insurers, and climate-finance programs willing to use these digital records for credit scoring, parametric insurance, and payments for ecosystem services. If that happens, a voice note sent from a hillside in Nariño could be enough to unlock working capital - turning WhatsApp chats into a financial identity for farmers who have never touched a CRM, but are now quietly generating clean, AI-ready datasets every day.

SOLYON Technologies

Under the radar in Medellín’s tech corridors, SOLYON Technologies is building the kind of AI infrastructure that banks, utilities, and governments quietly require but rarely advertise. The company focuses on cloud-native “sovereign” AI stacks designed so that sensitive data, identity, and compliance never leave approved boundaries.

Why regulated players care

For EPSs, financieras, and public entities across Colombia and the wider region, running machine learning in production is less a tooling problem and more a governance headache. SOLYON’s platform weaves compliance engines and identity systems directly into the MLOps layer, so every model, dataset, and API call can be traced, audited, and permissioned.

• Fine-grained identity and access controls baked into model deployment
• Policy engines that encode sector regulations into reusable templates
• Monitoring pipelines for data lineage, drift, and audit-ready logs

Backed by Medellín’s innovation hub Ruta N, SOLYON has already signed an institutional pilot to support high-stakes regulatory environments in the U.S. and Latin America. That cross-border proof point shows how sovereign-by-design infrastructure from Colombia can compete with heavier, global platforms.

Medellín as a sovereign AI lab

International investors increasingly describe Colombia as a launchpad into the broader region, with the Medellín-Bogotá axis offering competitive costs and strong engineering talent. Analyses such as Global Venturing’s overview of Colombia’s role in Latin America highlight how local startups are moving beyond applications into infrastructure. SOLYON fits that pattern: instead of one more chatbot, it is betting that regulated sectors will demand local control over AI pipelines, data residency, and identities.

What to watch next

The next few years will test how far a Medellín-born stack can scale. Growth hinges on alliances with Colombian telcos and cloud providers for hybrid deployments, on evolving AI regulation that rewards “compliance by design,” and on whether global clouds, cybersecurity vendors, or integrators such as IBM and Accenture eventually decide that acquiring sovereign infrastructure is faster than building it from scratch.

WeKall

On sales floors from Bogotá to Barranquilla, most calls still start the same way: a ringtone and a greeting. WeKall turns everything that happens after that into data. Born as a cloud telephony provider and now headquartered in Medellín’s SaaS corridor, the company has evolved into an AI-powered voice and sentiment analysis platform that helps sales and support teams treat every call as a measurable asset.

From dial tone to real-time insight

WeKall’s core stack ingests calls from distributed teams - whether an agent is in a WeWork in El Poblado or at home in Suba - and runs them through models that transcribe, categorize, and score interactions. After acquiring Sirenna AI, the startup significantly strengthened its ability to extract sales intelligence from those conversations and deliver coaching signals to managers while calls are still in progress.

• Automatic transcription and tagging of key intents, objections, and outcomes
• Sentiment and tone analysis to flag at-risk deals or unhappy customers
• Script-adherence checks for regulated industries that require call monitoring

Why Medellín is a natural launchpad

Medellín has quietly become a hub for B2B SaaS and customer-experience tools, supported by strong connectivity and nearshore talent, a trend highlighted in ecosystem profiles such as Built In’s overview of SaaS companies in Medellín. WeKall fits neatly into that story: it offers Latin American enterprises a way to modernize voice infrastructure without abandoning their existing telephony setups, while adding an AI layer tuned to Spanish and Portuguese conversation patterns.

Backed by corporate venture investors like LLYC Venturing, the company is increasingly relevant for sectors where calls are not just customer touchpoints but legal records - financial services, healthcare, and large retail among them. As Colombia continues to position itself as a regional tech launchpad, with venture analysts at Nomad Capitalist citing its growing startup and VC base, WeKall shows how an everyday channel like the phone can become a rich stream of AI-ready data.

Loto Latam

B2B sales rarely make headlines, but they keep Colombia’s construction sites, factories, and service contracts moving. Loto Latam focuses squarely on that under-automated world, using generative AI to streamline prospecting, follow-ups, and pipeline management for commercial teams. In late 2025, the company raised a US$1M (~4.0B COP) pre-seed round led by Fen Ventures, giving it the firepower to scale beyond its early Colombian base.

In many mid-sized companies, vendedores still juggle Excel sheets, email threads, and half-updated CRMs. Loto’s platform replaces that with AI agents that act like an extra commercial assistant working 24/7.

• Generate personalized outreach and follow-up messages tuned to each account
• Centralize communication across email, WhatsApp, and calls into one timeline
• Score leads and opportunities to highlight which deals deserve attention today
• Surface patterns in segments, geographies, and product lines to guide strategy

That promise has translated quickly into traction. Within months of launch, Loto scaled to more than 100 corporate clients - including real estate player Amarilo and KTM Powershop - and reported 0% churn over that early period, according to Cuantico VP’s analysis of Colombia’s top startups on LinkedIn. That stickiness hints that once sales teams embed Loto into their daily routine, they are reluctant to go back to manual workflows.

Looking ahead, Loto’s roadmap centers on scaling out of Colombia into Mexico, Chile, and Peru, where B2B sales still run on similar processes. Ecosystem observers at outlets like The Startup VC already frame Colombia as a launchpad into those markets. For engineers and product managers, Loto is a concrete example of how Spanish-first, region-aware AI tools can carve out defensible space before global CRM giants fully localize their own generative assistants.

NanoFreeze

Cold chains rarely make front-page news, but in Colombia they quietly drive the energy bills of supermarkets, tiendas de barrio, and pharma distributors. NanoFreeze sits at the intersection of bio-nanotechnology, industrial design, and AI monitoring to tackle that problem, redesigning freezers so they consume far less power without sacrificing performance.

AI-optimized “Natural Freezers”

The company’s flagship “Natural Freezers” combine advanced cooling materials with an AI layer that tracks temperature, door openings, compressor cycles, and performance anomalies in real time. According to Fast Company’s ranking of the most innovative Latin American companies, NanoFreeze has demonstrated up to 44% increases in cooling efficiency, a step change in a segment where single-digit gains are usually celebrated.

• Onboard sensors feeding AI models that learn store-specific usage patterns
• Dynamic control of cooling cycles to avoid unnecessary peaks
• Predictive maintenance alerts when performance drifts from the norm

In 2025, NanoFreeze ran a high-profile collaboration with Corona, rolling out energy-saving freezers across retail networks in Colombia and Mexico. That pilot turned abstract efficiency claims into utility bills and emissions baselines that operations teams could see, measure, and report.

From kilowatts to climate metrics

As large retailers and consumer-goods companies face pressure to report carbon reductions under frameworks promoted by initiatives like the Science Based Targets initiative, AI-instrumented equipment like NanoFreeze’s units becomes doubly valuable: it lowers the bill and produces auditable data. For ML and hardware engineers in Bogotá and Medellín, NanoFreeze is a concrete example of how applied AI can turn something as mundane as a freezer into climate-tech infrastructure, and how Colombian teams can ship that innovation across Latin America’s hot, power-hungry retail landscape.

Final Sip: Training Your AI Palate

The cupping table is quiet again. The last scores are scribbled down, the spoons rest on napkins, and ten codes are circled for next month’s chalkboard. You know, standing there with a lingering taste of panela and citrus, that dozens of other cups were just as complex. That’s what any “Top 10 Colombian AI startups” list really is: a snapshot of an ecosystem that refuses to fit on one menu.

Reading the flavor notes of ColombIA

Look back at the profiles and a pattern emerges. Colombia’s AI story leans into health, agtech, climate, and B2B workflows more than pure consumer playfulness. It reflects a country where clinics, farms, call centers, and logistics corridors are the main laboratories. Reports on sectors like software and IT in Cali and Valle del Cauca from agencies such as Invest Pacific show the same pattern: applied, operational AI built close to real problems, not just as demos.

The Bogotá-Medellín axis plays like two distinct origins on the same flight. Bogotá leans into regulated industries, enterprise SaaS, and fintech-adjacent tools. Medellín pushes infra, voice, and nearshore services, boosted by programs like Ruta N and dense BPO corridors. Both benefit from a steady flow of engineering talent and a cost structure that lets teams serve U.S. and European clients without Silicon Valley burn, a dynamic also noted in ecosystem overviews from outlets like Fintech News’ analysis of Colombia’s startup scene.

“The Colombian ecosystem shines for its resilience and its ability to turn challenges into opportunities. We’re not just enduring cycles of change - we’re growing through them, with innovation and purpose at the core.” - Miguel Vanegas Torres, Venture Partner, Impacta VC

For you, as an engineer, founder, or product person riding TransMilenio with arXiv tabs open, the real work starts after the list. Treat these ten startups as calibration points. Train your palate on what’s already brewing in health, campo, infra, and retail - and then look for the empty cups: public transport optimization, justice system backlogs, resilient housing, bilingual education. Somewhere between Chapinero and Laureles, those missing flavors are the ones you might roast next.

Frequently Asked Questions