Frequently asked questions

ZAIPEX is a Mexican artificial intelligence and data company. We exist because we got tired of seeing data projects fail — poorly implemented, disconnected from the business, or never reaching production.

We combine real technical depth with a business-results orientation. Here's what we do:

• Data engineering: connect, clean, and automate the data you already have
• AI Agents & RAG: intelligent automation and answers based on your own data
• Machine learning: predictive models deployed in production, not demos
• Dashboards & analytics: visualizations your team uses every day
• Data strategy: define what to build first and why
• Deep learning: high-complexity solutions — vision, NLP, anomaly detection

Additionally, through Zaipex Labs, our research lab, we build our own models, open source tools, and knowledge that feeds every project.

We don't sell tools. We don't do generic digital transformation. We implement AI and data that works, in production, with real users, with measurable impact.

Any company that wants to work with data — regardless of size, industry, or level of technical maturity.

• From startups to enterprises, in any sector
• If you have data and want to do something with it, we can help
• You don't need a technical team or prior AI experience to get started

If you have an idea, a problem, or just curiosity, the door is open. Reach out and let's talk.

Simple. No red tape.

• We talk: one or a few calls to understand your problem, your context, and what data you're working with
• Quick diagnostic: we look at what you have, its current state, and what's possible
• Clear proposal: what we'll do, what you'll get, and when — in under 5 business days
• NDA signed, we start building: we connect to your systems and get to work

No long-term commitments. We start with something concrete, deliver it, and if you're happy, we keep going.

Your data is yours. Always.

If we stop working together, we delete all copies of your information and give you a written certificate of destruction within 10 business days. Everything we built — dashboards, models, code — belongs to you. This is clear from the NDA we sign before touching any system.

No fine print, no lock-in. If tomorrow you want to take everything and run it with your own team, you can.

Pretty much anything that has data. Seriously.

• ERPs: Odoo, SAP, Oracle, Microsoft Dynamics, and more
• Databases: PostgreSQL, MySQL, SQL Server, BigQuery, Snowflake, etc.
• REST APIs, CSV/Excel, Google Sheets, Salesforce, HubSpot, and others
• Legacy systems — if it exports data somehow, we can connect it

The list goes on, but the rule is simple: if your system has data, we can connect to it. Ask us about yours.

We use production-proven tools, not whatever's trending.

• Data engineering: Spark, Airflow, dbt, Kafka, Delta Lake, and more
• Machine learning: scikit-learn, XGBoost, PyTorch, MLflow, etc.
• AI Agents: LangChain, OpenAI, Anthropic, Weaviate, Pinecone, and others
• Dashboards: Power BI, Looker, Metabase, BigQuery/Snowflake, etc.
• Deep learning: PyTorch, Hugging Face, YOLO, TensorRT, and more

We're not married to any tool. We always recommend the simplest thing that solves your problem — not the most expensive or the newest. If a spreadsheet does the job, we'll tell you.

Wherever you decide. We adapt.

• Ideally, we work directly on your own cloud infrastructure — so your data never leaves your environment
• If you don't have infrastructure, we use secure environments on AWS, GCP, or Azure with AES-256 encryption at rest and TLS 1.2+ in transit
• Need on-premise or hybrid? We do that too

Before we start, we put in writing where data lives, who has access, and under what conditions. No ambiguity.

Minimum access needed. Nothing more, nothing less.

• Every person on our team gets their own credentials — we never share access
• Multi-factor authentication (MFA) on everything
• All connections go through VPN, SSH, or IP restrictions
• You get a full log of who accessed what and when

When the project ends, we revoke all access and document it. Nothing stays open.

No. Period.

Your data is used only for your project. We don't train our models with client data — unless you explicitly ask us to, in writing. This is in our NDA and it's non-negotiable.

Zaipex Labs does its research with proprietary, public, or synthetic data. Never with yours.

We take security seriously. Here's what we apply on every project:

• TLS 1.2+ encryption in transit and AES-256 at rest — your data travels and is stored protected
• Role-based access control (RBAC) — everyone sees only what they need
• Multi-factor authentication (MFA) on all access points
• Real-time access monitoring and auditing
• If something happens, we notify you in under 24 hours — with an action plan included
• Our cloud providers hold ISO 27001 and SOC 2 Type II certifications

With a clear process at every stage:

• Before we start: we sign an NDA before seeing a single data table
• During the project: minimum access, encrypted and audited — everything is documented
• When it's done: we delete everything and give you written proof

It's not just policy — it's how we operate on every project. Full details at zaipex.ai/privacy.

Write to us at [email protected] and tell us what you need. Just include:

• Your name and the email you registered with
• A copy of your government-issued ID
• What right you want to exercise: access, correct, cancel, or limit the use of your data

We'll respond within 20 business days max, as required by law (LFPDPPP). In practice, we usually respond much sooner.

No. Never. We don't sell or share client data for commercial purposes. Plain and simple.

The only third parties that may have limited access are:

• Cloud providers (AWS, GCP, Azure) — because that's where projects run
• Authorized tax invoice providers (PAC) — for billing
• Authorities — only when there's a legal requirement

All of them are bound by confidentiality agreements as strict as the ones we sign with you.

It's our research lab. This is where we experiment, build our own models, publish open source tools, and generate the knowledge that feeds every implementation project.

It's not something you hire — it's what makes us different. Everything we learn through research ends up improving what we deliver to our clients.

Everything we publish under open source licenses (MIT, Apache 2.0, Creative Commons) is free to use. Models, datasets, tools — if it's published, it's yours to use.

Each publication clearly states its license. If you need something beyond that, write to [email protected] and let's talk.

We focus on what actually matters for companies in our region:

• AI in Spanish & LATAM: models that understand our context, not just translations from English
• Industry-applied AI: manufacturing, healthcare, retail — where generic models fall short
• Efficient models: quantization, distillation, pruning — AI that runs without a massive datacenter
• LLM evaluation: benchmarks, hallucination detection, metrics to know if a model actually works in production

Absolutely. We're open to working with researchers, universities, and technical teams who share our vision.

Send us an email at [email protected] with the subject "Zaipex Labs — Collaboration" and tell us what you have in mind. The worst that can happen is we end up having an interesting conversation.