Top 10 Companies Hiring AI Engineers in Colombia in 2026

Table of Contents

• Introduction: Colombia’s AI job map
• Mercado Libre
• Rappi
• Globant
• Bancolombia
• EPAM Systems
• Grupo Sura
• Ecopetrol
• McKinsey & Company
• Capgemini
• Avianca
• Conclusion: Using the map and next steps
• Frequently Asked Questions

Mercado Libre

For many AI engineers in Colombia, Mercado Libre is the bright red star over Bogotá. The company runs a major tech hub in the capital plus remote roles across the country, with AI/ML salaries that sit at the top of the local market: roughly 7-10M COP/month for juniors, 12-18M for mid-levels, 20-35M for seniors, and 38M+ COP/month for leads when RSUs and bonuses are included.

What you’ll build with AI

Mercado Libre organizes its data science and engineering teams into a unified “AI Flywheel” spanning marketplace, payments, and logistics. According to an earnings analysis reported by PYMNTS on Mercado Libre’s AI investments, this approach has helped support a ~45% regional revenue surge, so your models are wired directly into business KPIs.

• Real-time credit scoring and fraud detection for Mercado Pago across multiple countries.
• Large-scale recommendation systems that personalize search, feeds, and promotions for tens of millions of users.
• Logistics optimization for promised-delivery dates, routing, and last-mile capacity across LATAM.

Team culture and stack in Colombia

Day to day, you’ll work mostly in Python with deep learning frameworks like PyTorch and TensorFlow, deployed on AWS and tools such as SageMaker, while many backend services run in Go. Squads in Bogotá pair closely with counterparts in São Paulo, Buenos Aires, and Mexico City, so English is common in cross-border work even if local Slack is pure Spanglish.

The interview bar is high: expect serious coding interviews plus ML system-design rounds. Compensation benchmarks on Levels.fyi’s Mercado Libre Colombia page confirm that senior engineers can land packages well above standard Colombian developer ranges, which offsets Bogotá’s higher cost of living.

Best for you if…

Mercado Libre fits if you want to see your models hit production in multiple countries and don’t mind your success being tracked in dashboards: revenue lift, fraud losses avoided, minutes saved in last-mile delivery. Engineers who have already shipped recommendation engines, credit-risk models, or solid backend services in smaller Colombian startups - and who are comfortable owning both modeling and productionization - tend to thrive in this flywheel.

Rappi

For many Bogotá engineers, Rappi is the star pulsing over Chapinero and Quinta Camacho. The unicorn’s HQ is in the capital with strong teams in Medellín, and it leans into hybrid and remote options. Compensation is aggressive by local standards: roughly 6-8M COP/month for juniors, 10-15M for mid-levels, 18-32M for seniors, and 35M+ COP/month for leads once equity is factored in.

What you’ll build with AI

Rappi is essentially a real-time optimization engine draped in an orange app. AI engineers touch problems like:

• Dynamic pricing for delivery fees and RappiTurbo, balancing demand, distance, and rider supply.
• Hyper-local demand forecasting by barrio and time slot, sensitive to rain, payday, and fútbol matches.
• LLM-powered conversational support for shoppers and couriers in multiple Spanish dialects.

The data is messy - orders canceled mid-route, traffic jams on Avenida 80, sudden thunderstorms in Envigado - so your models must adapt quickly to a chaotic physical world.

Team culture and stack

Rappi runs a hub-and-spoke structure: a central research group with AI engineers embedded in product squads. The stack centers on Python (often with FastAPI), Hugging Face for NLP, scikit-learn, Airflow, and Snowflake. As noted in Expand’s AI engineer in Colombia report, roles here blur data science and engineering - you’re expected to own both models and production pipelines.

Best for you if…

This environment fits engineers who like high-speed, high-ambiguity work: streaming data, constant A/B tests, and on-call rotations. If you’ve built solid Python and DevOps foundations - perhaps through intensive programs like a 16-week back end and DevOps course - and now want to see them tested against Bogotá’s Friday-night rush, Rappi offers that pressure cooker at continental scale, without leaving Colombia.

Globant

If you want variety rather than a single product roadmap, Globant is often the first star you notice on the map. With sizable offices in Bogotá and Medellín plus distributed teams across the region, it runs one of Colombia’s most visible enterprise AI consulting practices. Compensation sits in the upper-middle of the market: roughly 6-9M COP/month for juniors, 11-16M for mid-levels, 18-30M for seniors, and 32M+ COP/month for leads, often paired with performance bonuses and international mobility.

What you’ll build with AI

Globant structures its data and ML work through an “AI Studio” model. Instead of focusing on one internal product, you ship solutions for Fortune 500 clients in North America and Europe. Projects commonly include:

• Generative AI copilots for customer support, internal knowledge bases, and developer productivity.
• Computer vision for manufacturing and retail, from defect detection to smart shelf analytics.
• AI-assisted test generation and code analysis that plug into clients’ CI/CD pipelines.

Tech stack and team culture

The core stack centers on Python, TensorFlow, and PyTorch running on AWS, Azure, or GCP, with MLOps tools like MLflow and Kubernetes becoming standard on larger programs. Because many engagements serve U.S. and EU customers, English is common in ceremonies and documentation, even though on-the-ground teams in Bogotá and Medellín are mostly Colombian and LATAM-based.

Globant appears prominently among the country’s leading AI vendors in listings like DesignRush’s top AI companies in Colombia, where firms are noted for their ability to combine engineering teams in Bogotá/Medellín with global delivery standards. That positioning shows up in the interview process: you’re tested not just on algorithms, but on whether you can jump from a retail recommender one quarter to an insurance document classifier the next.

Best for you if…

You want breadth and global exposure more than deep specialization in a single domain. Engineers who already have solid ML foundations - whether from university programs or focused Python/ML training - and are eager to experiment with gen AI, CV, and MLOps across multiple industries will find Globant a strong launchpad without leaving Colombia’s main tech hubs.

Bancolombia

In Medellín’s Ruta N district, Bancolombia’s glass towers are one of the brightest stars on Colombia’s AI map. It’s the country’s largest bank and a flagship employer for data and ML talent, with AI teams split between Medellín, Bogotá, and regional offices. Compensation is solidly corporate: around 5-7M COP/month for juniors, 8-13M for mid-levels, 15-25M for seniors, and 28M+ COP/month for leads, on top of bonuses that can reach 47% of annual salary for high performers.

What you’ll build with AI

The heart of Bancolombia’s ML work is its Analytics Excellence Center in Medellín. Unlike banks that keep AI in the “innovation lab,” this group ships models straight into core systems used daily by millions of Colombians. Typical problem spaces include:

• Fraud detection across cards, wire transfers, ATMs, and digital channels.
• Credit-risk scoring for consumers and SMEs, tightly governed by regulation.
• Personalization for mobile and agent channels, from product offers to next-best actions.

Team culture, stack, and stability

The stack blends Python and modern cloud data lakes with legacy tools like SAS, reflecting the bank’s gradual modernization. Senior engineers here earn in the 15-25M COP/month band, which lines up competitively against mid-senior engineering ranges of roughly 14-22M reported in regional studies such as LatAm engineering salary benchmarks. Glassdoor reviews consistently place Bancolombia near 4.6/5 stars for culture and benefits, emphasizing career paths and work-life balance.

Hybrid work is common, but many leadership roles remain anchored in Medellín, reinforcing the city’s identity as an “AI-for-industry” hub versus Bogotá’s higher-paying, more startup-heavy market. English matters for some vendor and regional projects, yet day-to-day collaboration is largely in Spanish.

Best for you if…

Bancolombia fits engineers who value impact and predictability over startup chaos. If you enjoy well-defined domains like fraud and credit, don’t mind wrestling with legacy systems and compliance, and want a long runway into roles such as lead data scientist, analytics manager, or product owner, this is one of the most stable AI homes you’ll find in Colombia - especially if you picture your life unfolding in Medellín rather than bouncing between cities.

EPAM Systems

On Colombia’s AI map, EPAM is the star that doesn’t sit in any one barrio. It’s remote-first across the country, with smaller hubs in Bogotá and Medellín, and it focuses almost entirely on building production-grade systems for U.S. and European clients. Compensation reflects a senior-heavy culture: mid-level engineers around 12-17M COP/month, seniors at 19-33M, and lead/principal roles from roughly 36M+ COP/month, often with USD-pegged components.

What you’ll build with AI

Rather than owning a single product, you’re the person companies call when their AI initiative needs to “grow up.” Typical engagements include:

• Designing end-to-end ML pipelines from ingestion in Spark through training and deployment.
• Implementing large-scale recommenders, pricing engines, or risk models that must withstand real traffic.
• Building shared data and MLOps platforms so client teams can ship models repeatedly, not just once.

Stack, culture, and nearshore angle

The stack blends Python, SparkML, XGBoost, and major deep learning frameworks with heavy use of CI/CD, feature stores, and observability tools. EPAM has been highlighted internationally as a “Most Loved Workplace,” and internally Colombia is known for a high concentration of staff engineers and architects who care deeply about software design, not just notebooks.

Most of your stakeholders are abroad. As reports on nearshore AI from firms like SuperStaff’s Colombia overview point out, time-zone alignment and cost competitiveness make Colombian teams attractive to U.S. clients, so nearly all collaboration is in English while your day-to-day life stays in Bogotá, Medellín, or your hometown.

Best for you if…

EPAM is a strong fit if you see yourself as an engineering-first ML practitioner: you enjoy architecture diagrams as much as model metrics, prefer remote work, and like jumping across industries while staying close to MLOps and platform problems. It’s especially appealing if you’re a seasoned software engineer moving deeper into AI and looking for peers who share that mindset.

Grupo Sura

On Medellín’s skyline, Grupo Sura is the red star that quietly spans both money and medicine. With headquarters in the Aburrá Valley and major teams in Bogotá, it hires AI talent into Seguros Sura (insurance) and Sura EPS (health services). Salary bands are competitive for a regulated sector: roughly 5.5-7.5M COP/month for juniors, 9-14M for mid-levels, 16-27M for seniors, and 30M+ COP/month for leads.

What you’ll build with AI

Few employers in Colombia combine actuarial science, clinical data, and ML the way Sura does. Typical work includes:

• Health-risk prediction to flag which patients may need preventive interventions or follow-up.
• Automated claims processing using NLP over medical notes, lab results, and invoices.
• Usage-based insurance models powered by telematics from vehicles, wearables, and IoT devices.

These aren’t “nice-to-have” pilots; they influence coverage decisions, pricing, and patient journeys across multiple countries.

Stack, culture, and ethics

The stack revolves around Python, cloud-native platforms on Azure and AWS, and specialized health-analytics tools. Teams are often matrixed: an AI engineer here might present a model to a squad that includes doctors, actuaries, and product managers. Ethical and regulatory constraints are front and center; interviews probe not just ML technique but how you reason about bias, explainability, and consent in health and insurance data.

Medellín’s universities, particularly EAFIT and Universidad de Antioquia, feed a steady stream of statisticians, actuaries, and data scientists into Sura, helping turn the valley into a regional “health-and-risk analytics corridor.” That specialization shows up in sector overviews such as Clutch’s Colombian AI firms directory, where healthtech and insurtech are singled out as growth niches.

Best for you if…

Grupo Sura is ideal if you’re drawn to AI for human well-being more than ad clicks or delivery ETAs. It suits data scientists moving deeper into healthcare or insurance, and ML engineers with strong NLP or time-series skills who want the trade-off of slightly lower upside than pure tech unicorns in exchange for mission, stability, and deep domain expertise in health and risk.

Ecopetrol

On Colombia’s AI map, Ecopetrol is the star pinned directly over refineries, pipelines, and seismic fields rather than co-working spaces. From Bogotá’s corporate offices to the ICP research center in Bucaramanga, its AI teams work at the core of the national energy grid. Salary bands reflect that responsibility: juniors around 6-8M COP/month, mid-levels at 10-16M, seniors between 18-30M, and leads from roughly 35M+ COP/month, plus sector benefits that are hard to match elsewhere.

What you’ll build with AI

Ecopetrol’s AI portfolio is deeply industrial. You’re more likely to be tuning a model on sensor streams than fine-tuning an LLM:

• Predictive maintenance for pumps, compressors, and pipelines using vibration, temperature, and pressure data.
• Seismic and geophysical analysis to assist exploration and quantify subsurface risk.
• Carbon-footprint optimization, modeling emissions and energy efficiency across complex operations.

Most problems revolve around time-series modeling, anomaly detection, and mathematical optimization, with safety and uptime as non-negotiable constraints.

Stack, culture, and physical reality

The typical stack blends Python for ML, MATLAB and domain-specific tools for signal processing, Azure cloud services, and industrial IoT protocols that connect SCADA systems and field sensors. Digital innovation teams are still relatively lean, which means more greenfield work and direct interaction with operations engineers. Interview loops emphasize statistics, control-theory intuition, and applied math as much as model-building. In national salary and hiring overviews like Alcor’s report on Colombian developer pay, energy and industrial roles consistently appear as premium tracks compared with generic software posts.

Best for you if…

Ecopetrol is a strong fit if you want your code to touch pipelines and power plants instead of dashboards alone. Engineers who started in mechanical, electrical, or petroleum fields and migrated into data/ML often thrive here, as do data scientists from manufacturing or utilities who want to move into higher-stakes industrial systems where every prediction can influence safety, production, and Colombia’s broader energy transition.

McKinsey & Company

If your compass points toward strategy and policy as much as code, McKinsey’s Bogotá hub is the red star worth circling. From a base in the city’s business district, its Agriculture & Nature Analytics and QuantumBlack teams work hybrid schedules and pay at the very top of the local market: mid-level roles around 15-22M COP/month, seniors at 25-40M, and leads at 45M+ COP/month, often with global-level bonuses.

What you’ll build with AI

Instead of optimizing click-through rates, you’re modeling how food, forests, and freight move across continents. Typical projects include:

• Geospatial ML for agriculture: yield forecasting, soil-moisture mapping, and crop disease detection using satellite and drone imagery.
• Sustainability analytics: deforestation risk, water usage, and carbon-credit modeling for governments and multinationals.
• Supply-chain optimization across food, mining, and logistics, with models feeding directly into board-level decisions.

Your work doesn’t just populate dashboards; it shapes national strategies and C-suite roadmaps.

Stack, culture, and expectations

The stack mixes Python, R, geospatial AI libraries, and proprietary QuantumBlack platforms. Roles are hybrid consultant-engineer: you’re expected to build robust models, interpret them for non-technical leaders, and contribute to slide decks and workshops. Reports on Colombia’s nearshore market, such as ProSource Technology’s 2026 IT staffing guide, highlight how global consultancies tap local AI talent for complex, high-value work - McKinsey is a prime example.

Best for you if…

McKinsey fits if you’re comfortable switching between Jupyter and PowerPoint, and you care about policy-scale and sector-scale impact. It’s a natural landing spot for people coming from academia, public policy, or economics who have solid ML foundations and want to see their models influence agriculture programs, climate commitments, or national logistics plans - while still living in Bogotá and plugging into the city’s broader AI community.

Capgemini

Capgemini is the bright blue logo you’ll spot in Bogotá’s El Chicó district, but most of its AI impact is behind the scenes: helping big retailers, banks, and telcos turn scattered models into something that actually runs in production. Salary ranges are competitive for consulting: juniors around 6-8M COP/month, mid-levels on 10-15M, seniors between 17-28M, and leads from 30M+ COP/month. In local engineering directories it’s often rated near 4.9/5 for culture and work-life balance.

What you’ll build with AI

Capgemini’s niche in Colombia is “industrial-strength” AI. Rather than inventing new algorithms, you make sure dozens of models behave in the wild:

• Designing robust MLOps pipelines so data scientists can deploy safely to cloud environments.
• Implementing monitoring, drift detection, and alerting for fleets of churn, pricing, and inventory models.
• Standardizing governance and documentation so regulators and internal auditors can trust AI decisions.

Stack and working style

The stack is cloud-agnostic, usually Python plus Azure or AWS, with reusable internal libraries for deployment, testing, and observability. Client teams are often in Europe, so English is common, but your day-to-day standups are still largely Colombian. Broader hiring scans like GoGloby’s Colombia tech overview highlight how firms such as Capgemini use Bogotá as a nearshore hub, mixing local talent with global delivery frameworks.

Best for you if…

Capgemini is a natural fit if you identify more as an ML platform engineer or DevOps-minded data scientist than as a research specialist. It’s ideal for software engineers moving into AI who love CI/CD and infrastructure, and for data scientists tired of notebooks that never make it to production. From Bogotá, you gain global enterprise experience in MLOps while still being paid in local currency and plugged into the city’s growing AI community.

Avianca

Avianca is the red star you notice every time you fly out of Bogotá. From its corporate HQ near the airport, the airline runs data and AI teams that sit right inside revenue management and operations. Compensation is solid corporate-tier for Colombia: juniors around 5-7M COP/month, mid-levels at 9-14M, seniors roughly 16-26M, and leads from about 30M+ COP/month, plus the usual airline perks.

What you’ll build with AI

Instead of optimizing clicks, you’re optimizing aircraft and tickets:

• Dynamic pricing for flights across routes, seasons, fare families, and booking windows.
• Crew and fleet assignment, solving large-scale scheduling problems under safety and labor constraints.
• Churn and loyalty modeling for LifeMiles and ancillary products like seats, bags, and insurance.

Stack, culture, and operations mindset

The stack is heavy on Python, Azure Databricks, and optimization libraries such as Gurobi, with classic ML for forecasting demand and no-show rates. Teams sit close to business owners; your stakeholders are revenue managers obsessing over load factor, yield, and on-time performance. Rather than deep neural nets, interviews emphasize optimization algorithms, experiment design, and how you’d validate a new pricing or scheduling model in production. Sector overviews like F6S’s list of AI companies in Colombia often single out travel and logistics as key local AI verticals, and Avianca is one of the few places where you tackle those problems at continental scale.

Best for you if…

Avianca is a strong home if you love operations research and applied math more than cutting-edge gen AI. It suits engineers from industrial, systems, or math backgrounds who enjoy turning messy constraints - crew rules, turnaround times, weather, demand spikes - into solvable models, and who want their work to show up in something very tangible: planes leaving on time and profitable routes connecting Bogotá, Medellín, Cali, and the wider region.

Conclusion: Using the map and next steps

You can hold the whole Top 10 in your hands like that tourist map on Carrera Séptima: ten bright stars - Mercado Libre, Rappi, Globant, Bancolombia, EPAM, Grupo Sura, Ecopetrol, McKinsey, Capgemini, Avianca - against a crowded backdrop of logos, barrios, and side streets you haven’t walked yet.

Reading the map, not worshipping it

The list is useful because Colombia’s AI market is noisy. Thousands of new IT graduates each year and a widening spread between junior and lead salaries make it hard to know where to start. Rankings compress that chaos into something you can scan on a bus ride from Suba to Chapinero or from Envigado to El Poblado - but they inevitably reflect someone else’s priorities: scale over lifestyle, salary over mentorship, prestige over stability.

Turn stars into your own criteria

The move is to flip the map around. Instead of asking “Which company is #1?”, ask “Which mix of city, domain, and culture actually fits how I want to live and work?” Bogotá might pull you with higher pay and the density of unicorns; Medellín might win with research-heavy roles and a slower daily rhythm. Nearshore consultancies, highlighted in market overviews such as Mismo’s analysis of AI careers in Colombia, offer another path: U.S.-facing work without leaving the country.

From here, your next steps look more like walking than ranking: visit meetups at local universities and tech hubs, talk to engineers inside these companies, ship a few portfolio projects, and apply with intention. Fold the list, pick one star as your starting point, and then give yourself permission to wander down an unmarked side street - that’s usually where the most interesting Colombian AI careers end up being built.

Frequently Asked Questions