AI: from strategy to production
AI can transform your business, if you invest in the right place and in the right way. We help you find where artificial intelligence creates the most value for your company, and then we take care of everything else: we design the solution, put it into production and run it with you. We move in stages, from a pilot to a fully operating service, so at every step you decide, based on real results, what to scale and how far.
Colpensiones
A compensation fund
A telecom company with operations across Latin America
Credifinanciera
JGB and Carvajal
Ingenio Mayagüez
Universidad Javeriana and UAO
Reybanpac, Ecuador
Six ways to bring AI into your company
AI agents
Agents that serve your customers, answer with reliable information from your documents and data, and handle requests for them. Includes semantic and image search for your products.
Dropi · Colpensiones · Reybanpac · Predizer TendersAI process automation
Long, repetitive tasks where a mistake is costly, done automatically: through the web or directly on your databases.
ADRES lookups · Certification paperwork · Activity-based (ABC) costing · Telecom tickets in 4 countries · Cámara de Comercio de CaliPredictive and prescriptive machine learning models
Forecasts of sales, production, risk and churn, recommendations on what to do, and the reasons behind each result.
JGB and Carvajal · Ingenio Mayagüez · Credifinanciera · Grupo Vardí · Smurfit KappaAI-assisted software development and SDD
Complete systems in weeks, built from a clear specification and backed by automated tests (spec-driven development).
CarLove Unicentro · Reybanpac · a compensation fundAI strategy and governance
Where to invest, what it returns and how to manage the risks, under the EU AI Act, the NIST AI RMF and the GDPR.
CoomevaAI training
Courses and workshops for executives and teams, with tools such as Claude and Gemini.
EAFIT · CIAT · Comfandi · Universidad Javeriana · INORCA · and 5 moreAI-assisted development, done properly
Using AI to write software is not asking a chatbot for code and seeing what comes out. That is vibecoding, and it leaves systems nobody can verify or maintain. We work with spec-driven development (SDD).
-
1
A specification you approve
Before writing code, we agree with you on what the system must do and how we will check it.
-
2
Automated tests
Every criterion in the specification is checked by an automated test that runs on every change.
-
3
A system your team can maintain
Documented, so your team, or your own agents, can maintain and extend it.
Predizer Tenders
Automatically finds the Latin American public tenders that fit your company. INORCA and MUBI Medical use it.
A pilot first, then the project
-
1
We decide where to invest
We identify with your team where AI creates the most value and how we will measure it.
-
2
We run a pilot
Before any large project, we test with your data to show that it works.
-
3
We build and deploy
We build the complete solution and put it into production.
-
4
We run and improve
We monitor the system in production and tune it over time.
They have trusted us
Dropi · Colpensiones · Comfenalco Valle · Coomeva · Credifinanciera · Reybanpac · INORCA · MUBI Medical · Grupo Vardí · Smurfit Kappa · Cartones América · Ingenio Mayagüez · Ingenio Pichichi · Universidad Javeriana · Universidad Autónoma de Occidente · Universidad de San Buenaventura · EAFIT · CIAT · Comfandi · Movistar · MAAS Transmilenio · PROQC International · INFODEC · Cámara de Comercio de Cali · CarLove Unicentro
“The work done by Predizer.ai […] is highly professional and adds immense value to the organization, with the assurance of basing decisions on scientifically developed arguments.”
“[…] His broad knowledge and experience in artificial intelligence, digital transformation and IT management are complemented by personal qualities and relationships that his clients, colleagues and teams value highly.”
“[…] In the projects we have worked on together, he has been a dedicated and highly committed professional, whose vast experience is matched by how up to date he stays as a curious and studious person. Highly recommended!”
“Thanks to Predizer.ai's knowledge, experience and strategy […], XOREX has implemented services and processes for our clients on artificial intelligence models, which make projects easier to deliver and add value to the service we provide our clients.”
“I had the opportunity to count on […] leadership and collaboration to extract, analyze and organize all the data of a specific credit product […] to build a predictive model for the default variable, keeping the number of variables to collect from each customer to a minimum […].”
You can trust us with your project and your data
We know what we do because we have been doing it for more than two decades, and we look after what you entrust to us as if it were our own.
More than 20 years in AI
Our consultants have applied artificial intelligence since 2002 and bring more than 30 years of experience leading technology at several companies. More than 25 organizations in six countries.
Systems still running
Colpensiones' system has been in production since 2021, and our forecasting models work on our clients' real data.
Your data, under your control
We can work inside your own infrastructure or cloud, so your data never leaves it. We only access what is needed, with permissions you grant and that are revoked when we finish. We never share our clients' data or code.
Responsible AI
We design your AI governance under the EU AI Act, the NIST AI RMF and the GDPR, as we did for Coomeva. Our models explain every result, and everything starts with a pilot to keep risk contained from day one.
The best models, through the safest route for your company
We work with Anthropic's Claude, Google's Gemini and many other models, and pick the best one for each task by quality and cost. We use them through the route you prefer: the provider's API, or your company's own cloud with AWS Bedrock, Google Cloud or Azure, in the region you choose. Always through enterprise services, where your data is not used to train the models.
Artificial intelligence with business experience
Predizer is an artificial intelligence consultancy based in Cali, Colombia, founded in 2019. We work in English and Spanish, with projects in Colombia, Ecuador, Peru, Venezuela, Chile and Taiwan.
We have spent more than 30 years leading technology at several companies and more than 20 applying artificial intelligence. That track record lets us speak the language of business and the technical one with equal ease. We trained at universities in Colombia and the United States, and university teaching taught us how to pass what we know on to your team.
Three ways to take the first step
Depending on where your company stands, we start with one of these three paths.
-
1
AI assessment
We review your processes with your team and come out with two or three prioritized use cases, with their expected return, ready for your budget.
-
2
A pilot on your data
An agent, an automation or a model running on your data, with metrics agreed from the start, before any large project.
-
3
Training for your team
Courses for executives and hands-on workshops for teams, with tools such as Claude and Gemini.
What people ask us most
Does my data have to leave my company?
It doesn't have to. We can work inside your own infrastructure or cloud, with permissions you grant and that are revoked when we finish. We never share our clients' data or code.
Which AI models do you use?
Anthropic's Claude, Google's Gemini and other models, and we pick the best one for each task by quality and cost. We use them through the provider's API or your company's own cloud (AWS Bedrock, Google Cloud or Azure), always through enterprise services where your data is not used to train the models.
What is spec-driven development, and why is it not vibecoding?
Before writing code, we agree with you on what the system must do and how we will check it. Then we build it with coding agents and automated tests that verify every criterion, and deliver it documented so your team can maintain it.
How long does a project take?
It depends on the scope. We have delivered automations in under a week, such as the ADRES member lookup, and complete systems in four to eight weeks.
How does a project start?
With a call to understand what you want to solve. Depending on the case, we follow with an assessment, a pilot on your data or training for your team.
Do you work outside Colombia, and in English?
Yes. We work in English and Spanish, with projects in Colombia, Ecuador, Peru, Venezuela, Chile and Taiwan.
Do you have a process that AI could do better?
Tell us what you want to solve. In a first call we will tell you whether AI is the right tool and what a pilot would look like. We reply within one business day.