You will get a custom AI agent trained on your own business knowledge base
Project details
A generic chatbot answers like it read your website once. This one answers like your best employee, because it is trained on your actual documents, SOPs and history.
I have shipped 36 AI agents to production for a marketing company, plus the agents that run guest communication and internal support for the hotels I operate. They run every day, with real users, and I fix them when they break.
WHAT I BUILD
• Ingestion: your PDFs, docs, sites and tickets chunked, embedded and stored in a vector database in your own account
• The agent: prompt, tools and guardrails tuned to your business, on Claude or OpenAI
• Tool calling: it reads and writes to your real systems, so it can look up an order, book a slot or open a ticket
• Channels: web widget, WhatsApp or Slack, sharing one conversation store
• Guardrails: source citations, refusal rules, human escalation, and a log of every answer it gave
WHAT MAKES IT HOLD UP
Retrieval has a relevance floor. Below it, the agent says it does not know instead of inventing something. Every tool call is logged and reversible.
You own the accounts and the knowledge base, and your team can update answers without a developer.
I have shipped 36 AI agents to production for a marketing company, plus the agents that run guest communication and internal support for the hotels I operate. They run every day, with real users, and I fix them when they break.
WHAT I BUILD
• Ingestion: your PDFs, docs, sites and tickets chunked, embedded and stored in a vector database in your own account
• The agent: prompt, tools and guardrails tuned to your business, on Claude or OpenAI
• Tool calling: it reads and writes to your real systems, so it can look up an order, book a slot or open a ticket
• Channels: web widget, WhatsApp or Slack, sharing one conversation store
• Guardrails: source citations, refusal rules, human escalation, and a log of every answer it gave
WHAT MAKES IT HOLD UP
Retrieval has a relevance floor. Below it, the agent says it does not know instead of inventing something. Every tool call is logged and reversible.
You own the accounts and the knowledge base, and your team can update answers without a developer.
AI Algorithms
Large Language Model, Transformer ModelAI Applications
AI Chatbot, Conversational AI, Natural Language UnderstandingAI Models
ChatGPT, GPT-4What's included
| Service Tiers |
Starter
$600
|
Standard
$1,600
|
Advanced
$3,400
|
|---|---|---|---|
| Delivery Time | 10 days | 21 days | 35 days |
Number of Revisions | 1 | 2 | 3 |
AI Model Integration | |||
Batch Normalization | - | - | - |
Database Integration | - | ||
Detailed Code Comments | |||
Image Upscaling | - | - | - |
MLOps | - | - | - |
Model Deployment | - | ||
Model Documentation | |||
Model Monitoring | - | - | - |
Model Testing & Optimization | - | ||
Model Tuning | - | - | |
Natural Language Processing | |||
NLP Tokenization | - | - | - |
Pre-Training | - | - | - |
Prompt Engineering | |||
Setup File | - | - | - |
Source Code | - |
About Daniel
AI Automation & Hotel Tech | Claude, n8n, Supabase, Vercel, Cloudbeds
Bogota, Colombia - 8:38 am local time
I lead operations and technology for two hotels (Puerto Rico and Colombia) and a restaurant, and I run the automation infrastructure for a marketing company. Every system I sell you, I built for a live operation first and fixed when it broke at 2am with real guests in the building. That is the difference between me and someone who only demos things.
WHAT I BUILD
AI agents and automation - Claude, Claude Code, OpenAI, n8n, Make, Zapier, GoHighLevel. Sales agents, internal chatbots, RAG knowledge bases, document generation, WhatsApp and CRM workflows. I run 30+ automations in production right now.
Full-stack apps - Supabase, Vercel, Next.js, Node.js, PostgreSQL. Financial dashboards, internal tools, client portals. Auth, RLS, edge functions, cron jobs, deployed and maintained.
Hotel and hospitality tech - Cloudbeds PMS end to end: setup, channel manager, booking engine, and OTA distribution across Booking, Expedia, Airbnb and Agoda without overbookings or rate parity issues. PriceLabs dynamic pricing. Direct booking strategy to cut OTA commission.
Operations and reporting - API integrations that turn raw data into automated P&L statements, occupancy reports and payroll. I built the systems that close the month for the properties I manage.
RESULTS I AM ACCOUNTABLE FOR
El Oasis Hotel (Aguada, Puerto Rico) - since I took over distribution, RevPAR is up 88% and occupancy went from 20.1% to 35.7%, with ADR up 5.7%. The gain came from distribution, not discounting.
Amara Beach Eco Hotel (Dibulla, Colombia) - RevPAR up 63% and occupancy from 12.9% to 29.1% in my first 12 months. WhatsApp is now the number two revenue channel and returns a higher ADR than the largest OTA on the account.
BlueHackers - Head of Operations. CRM pipelines, sales reporting, AI agents and the internal tooling the team uses daily.
HOW I WORK
I ask what the business actually needs before I touch a system. Most problems people bring me are not software problems, they are configuration problems. A hotel about to abandon its PMS almost never has a PMS problem.
Fluent in English and Spanish. Caribbean / US Eastern time zone.
Steps for completing your project
After purchasing the project, send requirements so Daniel can start the project.
Delivery time starts when Daniel receives requirements from you.
Daniel works on your project following the steps below.
Revisions may occur after the delivery date.
Scope and ingestion
We agree what it must answer, what it must refuse and what it may touch. Then I build the pipeline and validate retrieval on your own questions.
Build, test and hand over
Agent, tools and guardrails built, tested against a question set you write, deployed to your channel, documented and handed to your team.