of calendar time to build an AI physics tutor.
It is a pilot already running in production at a public university in Lima, not a prototype. Built with the SYLLPA factory.
AI agents execute. Human judgment directs. We build enterprise software with a factory designed around AI: more execution capacity, continuous validation and faster deliveries. Robust, secure and scalable from day one.
Let's talk about your project ↗SYLLPA is an AI-native agency in Peru. We use AI agents to build your software, and we analyze, case by case, where AI adds value inside the product and where a solution without AI is enough.
Specialized agents do the work under the direction of a SYLLPA person. An independent reviewer agent evaluates the results and a person approves before release.
Defines the structure and boundaries of the system before any code is written.
Designs how your information is stored and protected.
Builds the business logic and the connections between systems.
Builds the web screens your teams and customers use.
Creates the mobile apps.
Reviews access, permissions and validations.
Writes and runs the automated tests.
Orders the flows so they make sense at first glance.
Writes the text of every screen in one consistent tone.
It is a factory designed around AI.
The agents do the work. This system coordinates, keeps, controls and validates that work.
Coordinates specialized roles and tasks.
The context and decisions of each project are kept across sessions.
Permissions, policies and human oversight.
From analysis to deployment.
Tests, reviews and controls on every change.
A history of decisions and activities.
The controls for your project are agreed with you, in writing, before we build.
Every change is tested automatically before moving on. If a test fails, the process catches it before the change goes any further.
An independent reviewer agent, which did not take part in the build, reviews the result. Before release, a person approves it.
If someone tries to save them in the code, an automatic control blocks the change the moment it is saved.
Deleting data, releasing to production or sending anything to third parties needs a person to approve it.
Only what you can verify in our texts and in your contract.
Ownership of the source code and of what we build for you is agreed in writing in each contract.
Personal data we collect through our commercial channels is handled under Peruvian Law No. 29733 and our Data Processing Policy. Data processed on behalf of clients is governed by the contracts and documents that apply to each project.
See policies ↗Without your written authorization in a data processing agreement, we do not send personal data to an AI provider. Nor do we allow a provider to use your information to train models.
See the AI Use Policy ↗It is one of the security controls present in our cases; those of each project are agreed in writing.
of calendar time to build an AI physics tutor.
It is a pilot already running in production at a public university in Lima, not a prototype. Built with the SYLLPA factory.
They direct the AI-native methodology of SYLLPA.

More than 14 years in SAP, enterprise applications and digital transformation. For 10 years he coordinated SAP and applications at Confipetrol, a multinational services company for mining and energy in Peru and Chile: more than 20 applications under his responsibility, teams of up to 19 people across internal support, external consultants and a software factory, and operating cost reductions of up to 30% with RPA and applied AI. Before that he was an SAP security and authorizations consultant for Intercorp Retail group companies and an SAP ABAP consultant. Co-founder of SafeSchool, a company independent from SYLLPA.
View LinkedIn profile ↗
More than 15 years as a software engineer and architect. As CTO, he brings SYLLPA his experience in large-scale software architecture and validates the architecture of what we build. He built high-transaction-volume platforms for US companies: AI-driven automated due diligence at VisoTrust, financial wellness at Brightside, sales enablement at Brainshark, incentive programs for major corporations at Perks and the Broad Institute platform, as technical lead and architect at Encora, Belatrix and Avantica. He founded E-Factora, an electronic invoicing SaaS, and teaches Angular and .NET.
View LinkedIn profile ↗It is how SYLLPA builds software: specialized AI agents by role do the work under human direction, inside controls that apply to every change. The result is software with more tests, automated deployments and shorter delivery times.
No. Agents execute bounded tasks. People decide the scope, review the results and approve whatever cannot be undone, such as releasing to production or deleting data.
With automated tests on every change. An independent reviewer agent, which did not take part in the build, reviews the result, and before release a person approves it. Keys and passwords are kept out of the code, and a person decides anything irreversible. The controls for your project are agreed with you, in writing, before we build.
No. We analyze your case and put AI inside the product only where it adds value and financial return. Where a solution without AI is enough, we use that.
Tell us what you need and we will say, case by case, where AI adds value and where it is not needed.
Let's talk about your project ↗Applied AI, where it adds value and financial return.