Cloud Development

1. Actionable Intelligence: Provides users with the information they need to resolve questions and make decisions based on real interaction data.
2. Full Availability: Operates 24 hours a day, 365 days a year.
3. Native Bilingualism: Handles Spanish and English fluently in all interactions.
4. Simplified Maintenance: Its .md files allow easy updates of rules, criteria, and databases related to your company, products, or services.
5. Autonomy and Control: eSYNTAX develops the initial configuration of the files that train the AI. Afterwards, you can make updates yourself without requiring technical intervention.

1. Immediate 24/7 Responses: Information available at all times to resolve questions and support decision-making.
2. Intelligent Site Guidance: Helps users find information without friction.
3. Contextualized and Relevant Information: Responses aligned with the context to facilitate decision-making.
4. Efficient Resolution of Questions: Reduces the time required to obtain clear answers.
5. Personalized Assistance: Responses adapted to each user's intent and context.
6. Trust Through Clarity: Information is presented in a coherent, structured, and easy-to-understand way.
7. Unified Access to Knowledge: All company information is organized to be consulted intelligently and consistently.
8. Conversation Continuity: The experience maintains context between interactions to improve response accuracy.

1. AI Integration via PHP: The programming language used to connect the KyNEX chat with AI is PHP version 8.0 or higher.
2. Use of Google Gemini API: The KyNEX chat uses Google Gemini API (Flash and Flash Lite), a leading technology for long-context conversations. Its attention capacity and daily interaction volume are fully scalable, dynamically adjusting according to the client's subscribed plan (tier), enabling everything from efficient low-cost operations to massive high-demand flows without losing conversation memory.
3. Elastic Multi-Model Engine: Automatic "failover" system that alternates between the latest Google Gemini models (Flash and Flash Lite) to ensure the chat never goes down due to connection issues or having reached Gemini's maximum allowed limit.
4. Native Bilingualism (Spanish/English): Ability to detect and process both languages fluently, adapting to the user's market without the need for external translators.
5. Contextual Session Memory: Maintains a message history, enabling coherent and in-depth conversations where the AI remembers what has been said since the beginning.
6. Editable Semantic Knowledge Architecture: Data injection through external (.md) files, allowing to update information by type: sales, support, or knowledge base without modifying the source code.

1. Agents with Visual Identity: Dynamic rotation of up to 10 different agents (5 women and 5 men) with real names and photographs in each session, eliminating the coldness of a bot and generating human trust.
2. Persistent Navigation (Omnipresence): The chat accompanies the user throughout all sections of the website without closing and without losing the thread of the conversation when changing pages.
3. Inactivity Proactivity Algorithm: If the system detects that the user stops interacting, it launches an automatic follow-up question to regain attention or politely close the session.
4. Human Typing Simulation: "Typing indicators" with delays proportional to message length for a more natural and fluid experience.
5. Sound Notification on Message Reception: When an agent responds, the interface emits a sound to capture the user's attention.

1. Sales GPS (Smart Links): Ability to send direct links to relevant content within the client's website, guiding users toward relevant information.
2. Email Conversation History Delivery: Users can send themselves the complete conversation transcript by email for later reference.
3. Audit Trail for Business Owners: Every time a user sends the history, a copy is automatically delivered to the client's email so they know exactly what their visitors are asking.
4. Validated Lead Capture: Email structure validation system using Regex to ensure the prospect database is as real and useful as possible.

It is the brain and heart of KyNEX. It is the semantically structured and classified knowledge base that organizes your company's information as a connected system of concepts, products/services, processes, relationships, and rules/criteria.

This knowledge base is integrated into KyNEX via .md files, enabling the AI ​​to understand your business coherently and generate accurate, natural responses aligned with your actual business operations. Therefore, the quality of the responses depends directly on the Semantic Knowledge Architecture created by eSYNTAX about your company and product/service offerings.

1. Configuration Through an External .md File: Time zone, domain, and contact email settings so the system adapts to your company and website. If PHPMailer is used, all parameters for sending email are also defined.
2. Documented Logging (Timestamp): Automatic stamping of the exact date and time for each message and session start, providing full traceability.
3. Anti-Spam Shielding (Honeypot): Invisible protection against malicious bots through hidden verification fields that do not disturb human users.
4. Credential Security: Use of .env files to protect API keys and compliance with security standards to prevent leaks.
5. Rate Limiting: Time control between messages to prevent server saturation and API abuse.
6. Email Sending via PHPMailer or mail() Function: You can define whether the email sent to the user with the chat transcript is executed through PHPMailer or PHP's mail() function.

1. Full Corporate Branding: The design can be fully adapted to the client's visual elements, colors, and typography, making it appear as a native feature of the website.
2. The Chat Interface Has Two Instances: The first appears when the page loads in a compact format to obstruct website content as little as possible. The second appears when the chat window expands after clicking the first.
3. Advanced Content Formatting: Impeccable rendering of Markdown, bold text, lists, and automatic cleanup of long URLs for comfortable reading.
4. Session State Management: Use of cookies to remember whether the chat should remain open or collapsed according to user preference, improving UX.

NOTES:

1. The integration of KyNEX includes the following:
   • Uploading the ai-chat folder to the root directory of the client's hosting server. The folder contains the resources and functions.
   • Configuration of the .env file where the client parameters, as well as their API_KEY, are registered. The AI API token consumption cost is the client's responsibility and is paid directly to the provider (Gemini), ensuring transparency and the lowest usage-based pricing.
   • Identification of the language variable used by the client's website to manage its content.
   • Integration of the include instruction on the pages where required. The include is what executes KyNEX.
   • Configuration of the styles in the cascading style sheets (CSS) using the client's corporate colors.
2. The characters in the Information Limit include spaces and are considered per language (not collectively) and by type: sales, support, or knowledge base. For example, 3,000 characters are approximately equivalent to 1 page of a Word document. In other words, if a client requires a Semantic Knowledge Architecture of 9,500 characters, the cost would be $4,500 plus the proportional amount for the excess; which in this case would be $250.
   
   The total number of characters is counted based on the content of the Semantic Knowledge Architecture and not on the information provided by the client.
3. The prices:
   • The integration of KyNEX includes Spanish and/or English and by type: sales, support, or knowledge base.
   • The development of Semantic Knowledge Architecture is priced per language and by type: sales, support, or knowledge base.
   • Do not include VAT.
   • Are expressed in Mexican pesos.
   • Are one-time payments.
4. The Semantic Knowledge Architecture includes multiple modules depending on the type: sales, support, or knowledge base. These modules are adapted to each company's information, making their content unique for every implementation. This Architecture is developed in an .md file for each language.
5. Once the KyNEX chat is fully functional, the client operates under the pay-as-you-go model of the Google Gemini API through two linked platforms:
   1. Google AI Studio (Technical Control):
      • API_KEY: Generated here to activate the service.
      • Pay-as-you-go Plan: Allows commercial use by removing free tier restrictions after registering a credit card.
   2. Google Cloud Console (Financial Management):
      • Budgets and Alerts: Allows defining a monthly amount (e.g., $10 USD) and receiving automatic notifications when consumption reaches 50%, 90%, or 100% of that amount.
      • Quota Limits (Hard Limit): Sets a usage "ceiling" in the Quotas section. Once the limit is reached, the API stops responding, ensuring that the planned budget is never exceeded.