DATA PROCESSING FOR RAG & AI • LOCAL • PRIVATE

From the chaos of your mailbox to clean data for AI in .md

We process a specific email account (a department's mailbox), along with documents and your company website. All information is processed on our private, dedicated servers; we never access the client's infrastructure. We assist you remotely in data collection if you wish. We extract, clean and organize everything into .md (Markdown) files, designed to integrate with RAG, ChatGPT and NotebookLM. Private server processing · No external APIs · Total privacy

“We are not magicians. We just process data.”

View packages and pricing Request processing
AFR LOGIC - Data processing for RAG and AI: we extract, clean and organize emails, documents and websites into .md files with SHA-256 deduplication and private server processing.
100%
Private server processing
SHA-256
Cryptographic deduplication
.md
Gold standard for RAG
Zero
Cloud dependency
The Problem

The chaos of your mailbox drowns your AI

Years of information buried in a mailbox, documents everywhere, and an unstructured website.

📬

Information buried in mailboxes

Duplicate emails, endless threads, signatures, headers and footers that confuse AI. Centralizing company information becomes impossible when every mailbox is a dead-end maze.

📂

Data everywhere and unstructured

Word, Excel, PDF, loose emails, websites… nobody knows where anything is. And when you want to use that data in your knowledge base for AI, the HTML noise, duplicates and disorder make your RAG retrieve garbage instead of useful knowledge.

The result: you have a library without a catalog. Your AI can't learn from what it can't find.

The Solution

We extract, clean, organize and deliver

We process one specific email account (departmental mailbox) + documents + website. Nothing more, nothing less.

1

Extract

We extract text from the designated email account (PST/MBOX), associated documents (Word, Excel, PDF) and the company website.

2

Clean

We remove signatures, headers, footers, HTML noise and irrelevant metadata. We apply SHA-256 deduplication to eliminate duplicates.

3

Organize

We classify content by functional area (Customer Service, HR, Operations, Finance, etc.) according to the segmentation you provide.

4

Deliver

We return clean, organized .md (Markdown) files, compatible with RAG, ChatGPT and NotebookLM systems. All processing is done on our private servers.

One package = 1 email account (departmental mailbox) + related documents + company website. Need to process more than one account? Purchase a package for each one, as long as the mailboxes are provided separately.

Why it works

Two technical reasons that make the difference

It's not marketing. It's data engineering applied to AI.

📄

Markdown (.md): the gold standard for RAG

Converting to Markdown is not a whim. It's the format that retrieval systems (RAG) understand best: plain text, clear hierarchy, embedded metadata and zero noise. When you prepare data for RAG in .md, retrieval precision skyrockets because the model finds exactly what it needs, without distractions.

# .md example ready for RAG
## Conversation with client Acme Corp
Date: 2025-01-15 | Sender: juan@acme.com
[Clean content, no signatures, no HTML, no duplicates]
🔐

SHA-256 Deduplication: millimeter precision

SHA-256 deduplication is not optional: it's a necessity. Every email, every fragment, is reduced to a unique cryptographic hash. If two emails are identical (even if they come with different senders or dates), SHA-256 detects and eliminates them. This is vital for AI precision: fewer duplicates = fewer hallucinations. And everything runs 100% on our private servers, without your data leaving our controlled environment.

SHA-256: f7a3c9e2b1d4a8c6e5f0d9b8a7c6d5e4f3a2b1c0d9e8f7a6b5c4d3e2f1a0b9c8

Processing on our own private servers with no external APIs — Your data does not travel to third-party clouds. Everything is processed on our dedicated servers under NDA and AES-256 encryption. We do not retain data, we only deliver and delete.

Plans

Choose the package for your functional area

Each package covers 1 specific email account (departmental mailbox) + associated documents + website. All prices include VAT. Invoicing in MXN at the day's Banxico exchange rate.

⏱️ Delivery times: Depend on volume. Google download is limited to 50,000 emails per day. The indicated timelines are approximate and begin upon receipt of data.

Starter
$900
USD · ~$18,000 MXN
Ideal for validation and small projects
  • ✓ 1 email account (up to 5,000 emails)
  • ✓ 20 associated documents
  • ✓ Website (200 pages)
  • ✓ Basic extraction and cleaning
  • ✓ Delivery in .md
Time: 1-2 days
Choose Starter
Most popular
Growth
$4,500
USD · ~$90,000 MXN
SMEs with medium volumes
  • ✓ 1 email account (up to 50,000)
  • ✓ 50 associated documents
  • ✓ Website (200 pages)
  • ✓ SHA-256 Deduplication
  • ✓ Delivery in organized .md
Time: 3-4 days
Choose Growth
Pro
$9,000
USD · ~$180,000 MXN
Consolidated companies, agencies
  • ✓ 1 email account (up to 150,000)
  • ✓ 100 associated documents
  • ✓ Website (300 pages)
  • ✓ SHA-256 Deduplication
  • ✓ Area classification
  • ✓ Delivery in organized .md
Time: 7-10 days
Choose Pro
Enterprise
$15,000+
USD · from $300,000 MXN
Large companies, government
  • ✓ 1 email account (up to 500,000+)
  • ✓ 500+ associated documents
  • ✓ Website (500+ pages)
  • ✓ SHA-256 Deduplication
  • ✓ Advanced area classification
  • ✓ Processing on private dedicated servers with Enterprise capacity
  • ✓ Data consulting included
Time: 3-4 weeks
Choose Enterprise
Custom
Custom quote
VAT included
Volumes over 500,000 emails or specific needs
  • ✓ We process the volume you need
  • ✓ Flexible timelines to be agreed
  • ✓ Complex classification or multiple accounts
  • ✓ SHA-256 Deduplication and deep cleaning
  • ✓ 100% local processing
  • ✓ Data consulting and support
Request quote
Success story

From a chaotic mailbox to 47 .md files

📊

15 years of history in a single mailbox

A services company came to us with email chaos accumulated over 15 years: 300,000 emails in one account, 92 additional information sources between documents and systems, and decades of information buried in mailboxes.

We extract, clean and organize all the content of that mailbox. We remove duplicates with SHA-256 deduplication, remove signatures and HTML noise, and classify by areas: customer service, personnel, operations and finance. We deliver 47 .md files perfectly structured.

.md ready for RAG SHA-256 Deduplication Local processing
300K
emails processed
→ 47
.md files
~2
weeks of work
The client now uses that base for their internal assistant. The AI finds answers without getting lost in the noise.
Clear limits

What we DON'T do

So there are no surprises. We are text data specialists, not everything else.

❌

Process images, audio or video

We only extract URLs, we don't do OCR or multimedia analysis.

❌

OCR of scanned PDFs

We work with extractable digital text, not images.

❌

Train AI models

We prepare your data. You use it in your RAG, ChatGPT or NotebookLM system.

❌

Store client data

We process, deliver and delete everything from our systems.

❌

Generate metadata or summaries

AI already does that on its own. We just prepare the clean text.

❌

Create playbooks or manuals

That's consulting, not processing. We don't offer that service.

Need something else? Talk to us and we'll tell you if we can help.

Why us?

Different approach, real results

Aspect Other approaches Our approach
Scope Process a single URL or document Process a complete email account + docs + web
Deliverable Raw data, unorganized Clean, organized data in .md
Time Per unit or per hour Per complete project (fixed price)
Infrastructure In the cloud (external APIs) Private dedicated servers
Focus For developers For business owners (clear results)

Commercial conditions

💵

Payments and currency

50% upfront / 50% upon delivery. All prices include VAT. Invoicing in MXN at the day's Banxico exchange rate.

📅

Operating days

Available every day. The timeline starts when the upfront payment is made and the data has been received.

🔒

Confidentiality

Processing on private servers. We do not retain data. NDA available upon request.

🔄

Cancellation and refunds

  • Upfront: 50% of total cost.
  • Cancellation before sending files: 100% refundable.
  • Cancellation after sending files: 20% retained for work done. The remaining 80% is refunded.
  • Cancellation by AFR LOGIC: 100% refundable.
📦

More than one email account?

Purchase one package per account, as long as the mailboxes are provided separately by the client. If the content is mixed, additional classification service is required (quoted separately).

Fresh knowledge

Latest blog articles

Guides and reflections on data processing, RAG and preparing information for AI.

Agentes de IA para correos: cómo automatizar respuestas con tu propio estilo | AFR LOGIC
📅 29/08/2026

Agentes de IA para correos: cómo automatizar respuestas con tu propio estilo | AFR LOGIC

Read article →
Auditoría de Ciberseguridad: El pilar invisible del procesamiento de datos para IA | AFR LOGIC
📅 29/08/2026

Auditoría de Ciberseguridad: El pilar invisible del procesamiento de datos para IA | AFR LOGIC

Read article →
Comercio, ERP y POS: El triángulo de los datos en el retail moderno | AFR LOGIC
📅 29/08/2026

Comercio, ERP y POS: El triángulo de los datos en el retail moderno | AFR LOGIC

Read article →
View all articles →
Trust

What our clients say

Real results, companies that trusted us. (All testimonials are anonymous due to NDA).

★★★★★

"We had 15 years of accumulated emails in our customer service mailbox. We honestly didn't know where to start. They turned everything into organized .md files by area. Today our internal assistant finds answers in seconds. Incredible."

Director of Operations

Service company, logistics sector

NDA protected
★★★★★

"What worried us most was security. We didn't want our information to go to the cloud. The process was 100% local, they sent us the files already ready. Also, we had many duplicate emails that were ruining our searches. They deduplicated and everything was immaculate."

CTO

Fintech, financial sector

NDA protected
★★★★★

"We thought it was expensive because it was the first time and we didn't know the market. But they were patient with us, explained everything, and in the end the service saved us a lot of time. We no longer have to waste hours looking for old emails. It was completely worth it."

Purchasing Manager

Manufacturing company, industrial sector

NDA protected
★★★★★

"We are a large company and we were afraid that someone would see our information. They work under NDA and with local processing, that's what convinced us. Not only that, but the end result was better than we expected. The .md files are perfectly organized and ready to use with our AI. Honestly, we couldn't be happier."

CEO

Educational technology company

NDA protected
Start today

Ready to prepare your data for AI?

You have years of information accumulated in a department's mailbox, documents, and your website. We deliver everything clean, organized, and in .md. Your AI does the rest.

No impossible promises. Just well-done work. SHA-256 Deduplication · Private server processing · .md ready for RAG

Request Processing Service Data Consulting

📬 Do you have PST or MBOX? We convert to Markdown with signature cleaning, HTML noise removal, and SHA-256 deduplication.