You will get an AI assistant that reads your inbox and calendar and briefs you daily
Project details
I built this for my own company and I use it every morning. It reads my email, messages, calendar and open work, then tells me what needs an answer, what I'm waiting on, and what matters today. Urgent things push to me instead of sitting there waiting to be found.
This is that, built for you.
The design matters more than the feature list. Rules that have to fire on every single run live in always-applied instructions, never in retrieval, because retrieval is probabilistic and policy can't be. Reference material lives in skills that load when they're relevant. Nothing irreversible runs without your approval. Every automated action gets logged.
I know its failure modes because I live in them.
Scope at this price: two sources, normally email and calendar, and one delivery channel. If you want your CRM, your project tracker and your phone in the loop too, that's a bigger build and I'll quote it honestly rather than squeeze it in here and do it badly.
What I need from you: OAuth access to the accounts involved and one conversation about what your morning actually looks like. Not what you wish it looked like.
Before you buy, click the portfolio on my profile. Two working demos.
This is that, built for you.
The design matters more than the feature list. Rules that have to fire on every single run live in always-applied instructions, never in retrieval, because retrieval is probabilistic and policy can't be. Reference material lives in skills that load when they're relevant. Nothing irreversible runs without your approval. Every automated action gets logged.
I know its failure modes because I live in them.
Scope at this price: two sources, normally email and calendar, and one delivery channel. If you want your CRM, your project tracker and your phone in the loop too, that's a bigger build and I'll quote it honestly rather than squeeze it in here and do it badly.
What I need from you: OAuth access to the accounts involved and one conversation about what your morning actually looks like. Not what you wish it looked like.
Before you buy, click the portfolio on my profile. Two working demos.
AI Algorithms
Large Language ModelAI Applications
AIOps, Conversational AI, Natural Language UnderstandingAI Models
ChatGPT, GPT-4What's included $1,200
These options are included with the project scope.
$1,200
- Delivery Time 14 days
- Number of Revisions 2
- AI Model Integration
- Model Deployment
- Model Documentation
- Model Monitoring
- Prompt Engineering
- Source Code
About Jose
AI Automation Engineer | n8n, APIs, AI Agents | 8 Yrs Full-Stack
Kissimmee, United States - 2:33 pm local time
I've spent 8 years building the systems businesses run on: CRMs, intake pipelines, internal tools, integrations. Now I connect AI to those systems so the work that eats your team's hours happens on its own.
What I build:
- Workflow automation with n8n, Make, and Zapier
- AI agents and chatbots wired into real processes (OpenAI and Claude APIs)
- CRM cleanup, migrations, and integrations
- Internal tools and dashboards
- Document and intake automation, in English and Spanish
Recent builds:
- A bilingual AI intake agent that qualifies leads, scores them, and books consultations on its own
- A Telegram AI assistant that runs daily operations from a plain-text knowledge base
- A native iOS app with paid subscriptions, live on the App Store (Swift, CloudKit, Watch app)
I write the code myself. No handoffs, no project managers. You talk to the person building your system.
If you have a workflow that leaks time or leads, message me. I'll tell you whether it's worth automating and what it would take. If it's not a fit, I'll say so.
Steps for completing your project
After purchasing the project, send requirements so Jose can start the project.
Delivery time starts when Jose receives requirements from you.
Jose works on your project following the steps below.
Revisions may occur after the delivery date.
Connect the sources
Your email and calendar connect through OAuth. Nothing gets copied out of the systems it already lives in.
Write the rules, then the briefing
Rules that must fire every run go in always-applied instructions, not retrieval. Then the briefing gets tuned against your real inbox until it reads like something worth opening.