We build custom AI, data and cloud solutions.
We design and build agents, modernize systems, and integrate data and cloud services with your existing technology. We agree on scope, acceptance criteria and costs, validate through a pilot, and support operations and handover to your team.
Your systems, data and rules shape the project.
- 01
Specific integrations and business rules
Your operation may need connections to core systems, data and rules that require configuration or custom development. The project must account for those dependencies.
- 02
Systems that need documentation
Business logic and dependencies accumulated in legacy systems can make changes difficult. Understanding them and checking them against tests helps plan modernization.
- 03
Requirements for production
Beyond testing a function, the project must define how it will be integrated, monitored and maintained. Data, architecture, security and support are part of that preparation.
Five capabilities to build on your existing technology.
We select and combine these capabilities according to the problem, the systems involved and the acceptance criteria. A project can include one or several areas of engineering, with deliverables and responsibilities defined from the start.
- 1
Custom agents and automation
Agents and copilots that consult sources and carry out tasks across multiple steps in your systems. We define tools, scope and points of human intervention for each process.
- 2
AI-assisted modernization
Analysis of legacy code, dependencies and security findings, with documentation and diagrams to support technical review. We design refactoring and integration through APIs or microservices when the use case calls for them.
- 3
Data and models
We integrate and prepare data for analytics and predictive models, including propensity, risk and campaign optimization. Model deployment and monitoring are included according to the project scope.
- 4
Cloud architecture and operations
Architecture, migration, continuity, observability and cost management. We assess hybrid or multiple-provider environments according to requirements and validated configurations.
- 5
AI security and governance
We define controls for prompts, sensitive data and agent use, with traceability of outputs and clear operating responsibilities. Governance is adapted to the environments included in the project.
Clear responsibilities for delivery, operations and handover.
- Integration with your systems
- Define connections to core systems, data sources and existing interfaces within the agreed scope.
- Quality and acceptance
- Agree on criteria before building and verify them with data, code and scenarios representative of your operation.
- Scope and total cost
- Evaluate deliverables and development and operating costs, with the assumptions and dependencies needed to compare alternatives.
- Operations and support
- A dedicated team, agreed service levels and 24/7 coverage for critical incidents.
- Continuity and handover
- Code, infrastructure as code, documentation and component rights defined in the contract, with knowledge transfer and a supported practical exercise for your team.
Projects in integration, modernization, data and AI.
Amarello implementations since 2008, across industries
Google Cloud Services Partner of the Year
share of response time consumed by transactional systems at Actinver
Team certified in Google Cloud and AWS. Each project's environment is validated against its components and requirements.
AI-assisted modernization · Insurance
A leading insurer in Mexico
Analysis and documentation of about 941,000 lines of COBOL, with code logic checked against QA tests. The analysis platform was deployed in the client's cloud. AI operating costs fell by more than 60% compared with the project's initial approach.
Read the full case studyDocument automation · Insurance
An insurer within a financial group with a national presence
Classification, extraction and validation of claims files against business rules, integrated into the operation. The analyst reviews exceptions and makes the decision, with each step recorded.
Integration and cloud · Banking
Actinver
An API and container layer connected existing systems. According to the Google Cloud case study, the share of response time consumed by transactional systems changed from 90% to 10%.
Read the Google Cloud case studyData and automation · Media
El Universal
Data integration and automation with Google Cloud. The published case reports up to 40% time savings on tasks previously performed manually.
Read the Google Cloud case study (Spanish)Predictive models · Airlines
Aeroméxico
Machine learning models for digital campaign optimization. Amarello's published case reports 35.3x ROAS, a 146% increase in conversion rate and US$170,000 in incremental revenue in the first month.
Read the full case studyCompanies Amarello has worked with
Banking · Insurance · Media · Airlines
We move phase by phase, with verified results.
We define the problem, alternatives and acceptance criteria before building. We test with your data or code, validate integration, and agree on how the solution will be operated and handed over.
- 1
Scope and alternatives
- What we do
- Define the problem, solution options, deliverables, acceptance criteria, and estimated development and operating costs.
- Your team's role
- Business and technical owners, agreed access, and a sample of data or code.
- Criteria to proceed
- Agreed scope, selected approach, cost assumptions and evaluation criteria.
- 2
Design and testing
- What we do
- Design the architecture and test critical components with representative data or code. Assess integration and feasibility against the defined criteria.
- Your team's role
- A technical counterpart and validation of rules, dependencies and results.
- Criteria to proceed
- Reviewed design, documented test results and necessary adjustments identified.
- 3
Operational pilot
- What we do
- Test with users, volumes and scenarios representative of the operation, within the pilot scope.
- Your team's role
- Users, authorized data and deployment windows coordinated with the responsible teams.
- Criteria to proceed
- Evaluated integration and performance, with an agreed decision on adjustments and production rollout.
- 4
Production and handover
- What we do
- Deploy in the validated environment, agree on support and deliver code, infrastructure as code and documentation according to the contract. Support your team as it carries out a significant change.
- Your team's role
- Technical and operations owners participating in knowledge transfer and the supported exercise.
- Criteria to proceed
- Reviewed acceptance criteria, defined support and handover verified with your team.
Architecture, integrations, security and deployment.
Deployment environments
Google Cloud Platform (GCP) is our primary runtime environment. The team also holds AWS certifications. Before defining deployment, we review the project's components, integrations and requirements, identifying what is validated in the proposed environment and what needs additional testing.
Modernization
Automated analysis of code, dependencies and security findings, with references to CVE and OWASP where applicable. Documentation, diagrams and a phased refactoring plan support technical review. AI-generated results are checked against the project's code and tests.
Data
Lakehouse architecture, data integration and transformation pipelines, and MLOps for deploying and monitoring models in production. Components and metrics are selected according to the sources, volumes and purpose of each model.
Agents
Agents with defined tools and tasks, integrated with the sources and systems included in the project. The workflow provides traceability of outputs and human checkpoints for decisions that require them.
Security and governance
Prompt governance, controls for sensitive data and traceability of agent use. We define responsibilities and operating rules for the included environments, aligned with the client's requirements.
Deliverables and handover
The contract defines code, infrastructure as code, documentation, component rights and handover to your team. Before closing the production delivery phase, your team carries out a significant change with our support to verify the handover.
What we get asked before we start.
When does custom development make sense?
When integration requirements, business rules or expected behavior need specific engineering work. At kickoff, we assess whether to build, adapt an existing tool or keep the current process. The decision is based on scope, cost and acceptance criteria.
Does the service include assessment and design?
Yes. The work can cover assessment, architecture, development, integration and operations, according to the agreed scope. We define deliverables for each phase and review alternatives before committing to development.
Does modernization mean replacing the entire system?
The scope is defined after reviewing the code, dependencies and business needs. It may include documentation, analysis, refactoring or phased integration. We agree on what stays, what changes and how the expected behavior will be validated.
Can you work with AWS, Azure or our own infrastructure?
Google Cloud is our primary environment, and the team also holds AWS certifications. For any environment, including Azure or your own infrastructure, we review which components and integrations have been validated and which additional tests are needed. Feasibility and scope are confirmed before committing to deployment.
Who designs and delivers the project?
Senior professionals participate in each phase with defined responsibilities. Before presenting the proposal, we introduce the team, its experience and its role in design, development and operations.
How are cost, timelines and deliverables defined?
We start with scope, dependencies, integrations and acceptance criteria. We estimate development and operations, define deliverables by phase, and agree on a schedule based on available access and resources. Assumptions and adjustments are reviewed with your team throughout the project.
Who handles a critical incident outside business hours?
A response team provides 24/7 coverage for critical incidents. Mitigation and restoration times are defined by priority, with four escalation levels up to the CTO, according to the agreed support terms.
How does Amarello support continuity and handover?
The contract defines deliverables, component rights, documentation and handover. Before closing the production delivery phase, your team carries out a significant change with our support. This exercise verifies that the team can work with the solution and the documentation provided.
We also solve
- Cloud, data and governanceCloud architecture, migration, modernization and operations, with data management, observability and governance aligned with project requirements.See how we work on it
- AI document processingClassify documents, extract information and validate data against your business rules. Bring together supporting evidence so your team can review exceptions and make decisions.See how we work on it
Tell us what you need to build, integrate or modernize.
We review your process, systems and operating constraints. Together, we define an initial deliverable, how to evaluate it and what you need to operate and maintain it.