Ricardo Munoz

Portrait of Ricardo Munoz
Add your portrait

I design AI agents and workflow automation around how people actually think and decide, so routine work runs itself and people keep the judgment calls.

Follow the signal to my work
  • Open to work
  • California

Most automation copies the steps. I design for the thinking behind them.

My cognitive science background shapes how I build. I map how a team actually makes decisions, then design agents, automations and human-in-the-loop checkpoints that fit that process, not the other way around.

Selected work

Open a project for the full breakdown: how it works, what it changed and the stack behind it.

  1. AI agent, customer operations2025

    Support inbox triage agent

    About 70% of tickets triaged automatically

    AI agent, customer operations2025

    Support inbox triage agent

    An LLM agent that classifies inbound email by intent and urgency, drafts grounded replies from the help-center knowledge base, and routes edge cases to a human-in-the-loop review queue.

    Outcome

    About 70% of tickets triaged automatically. First-response time dropped from 6 hours to 15 minutes.

    Diagram: messages flow from the inbox to a triage agent, which labels, drafts a reply or escalates to a person.
    Skills
    • Intent classification
    • Prompt engineering
    • Human-in-the-loop design
    • LLM evaluation
    Tools
    • Python
    • OpenAI API
    • Gmail API
    • PostgreSQL

  2. Workflow automation, revenue operations2025

    Lead enrichment and routing pipeline

    Speed-to-lead cut from a day to under 2 minutes

    Workflow automation, revenue operations2025

    Lead enrichment and routing pipeline

    An event-driven workflow that enriches inbound sign-ups with firmographic data, scores ICP fit with an LLM, writes the record to the CRM and alerts the owning rep, with retries and deduplication built in.

    Outcome

    Speed-to-lead cut from about 24 hours to under 2 minutes.

    Diagram: a sign-up form is enriched, scored by an AI lead scorer, then written to the CRM and sent as a Slack alert.
    Skills
    • Workflow orchestration
    • Data enrichment
    • Lead scoring
    • Error handling and retries
    Tools
    • n8n
    • HubSpot API
    • Clearbit
    • Slack API

  3. Agentic RAG, knowledge management2024

    Internal knowledge assistant

    Hour-long lookups answered in under a minute

    Agentic RAG, knowledge management2024

    Internal knowledge assistant

    Retrieval-augmented generation over 2,000+ internal documents. The agent plans sub-queries, retrieves and re-ranks passages, and answers with inline citations for every claim.

    Outcome

    Research lookups went from about an hour to under a minute, with each answer traceable to its source.

    Diagram: a question goes to a research agent that loops with a document library before returning a cited answer.
    Skills
    • Retrieval-augmented generation
    • Agent orchestration
    • Embeddings and re-ranking
    • Grounding and citations
    Tools
    • LangGraph
    • pgvector
    • TypeScript
    • Anthropic API

  4. Systems integration, IT operations2024

    Employee onboarding automation

    Provisioning per hire from 3 hours to 10 minutes

    Systems integration, IT operations2024

    Employee onboarding automation

    An HRIS-triggered workflow that provisions accounts across five SaaS tools, applies role-based access and posts a day-one checklist, with an audit log for IT.

    Outcome

    Provisioning per hire went from about 3 hours to 10 minutes, with no missed accounts.

    Diagram: a new hire event triggers an orchestrator that provisions Google, Slack, GitHub, Notion and payroll.
    Skills
    • Systems integration
    • Role-based access control
    • Process mapping
    • Audit logging
    Tools
    • Make
    • Google Workspace Admin API
    • Slack API
    • Notion API
Portrait of Ricardo Munoz
Add your portrait

Ricardo Munoz

I build AI agents and workflow automation. I design around how people decide, so routine work runs itself.

  • Open to work
  • California
  • B.S. Cognitive Science

About

Portrait of Ricardo Munoz
Add your portrait Save a photo as assets/portrait.jpg. A 4:5 crop of at least 800 by 1000 pixels works best.

Cognitive science graduateCalifornia

Ricardo Munoz

I’m Ricardo, a 23-year-old builder with a B.S. in Cognitive Science. I studied memory, attention and decision-making under pressure, and now apply that to LLM agents, workflow automation and the integrations that connect them, so teams spend less time on routine work and more on the decisions that need judgment.

Skills and experience

AI and agents

Automation and integration

Coding

Cognitive science

Experience

  1. Cognitive Science Lab, University name, 2021–2023

    Undergraduate research assistant

    Attention and decision-making experiments, with Python analysis pipelines.

  2. Company name, 2023–2024

    Operations systems associate

    Internal tooling and integrations across CRM, HRIS and ticketing.

  3. Independent, 2024–now

    AI automation engineer

    LLM agents and automations for ops, sales and support teams.

Have a process that should run itself?

Tell me about it on LinkedIn, or read through my code on GitHub.