Executive Assistant + AI Automations Engineer

I run the operations behind founders, and build the AI that scales how they work.

Executive Assistant + AI Automations Engineer. Equal parts sharp executor and curious builder, with zero dropped balls.

View Resume (PDF) ► Voice: Hardest Project ► Voice: What Drives Me LinkedIn
60%
Less manual invoice processing
85-92%
AI knowledge-base accuracy
5+ hrs
Saved per week, per workflow
95%
Document extraction accuracy
About

Two skills that rarely live in the same person

I work at the intersection of running an executive's day-to-day with zero dropped balls, and building the AI agents and workflows that make that operation scale.

I have supported senior executives and founders across calendar, inbox, travel, and relationships, while designing automation that removes the repetitive work entirely. What drives me is leverage: taking a messy, manual process and turning it into a reliable system that hands time back to the person I support.

A note on the work shown here.
Out of respect for client confidentiality, this portfolio shows my methods and the systems I build, using templates I created and sanitized case studies, never a client's private data.
Executive Assistant

The systems that keep a founder's day running

Each capability below is backed by a real template I built. They double as the operating system behind the executive I support.

📅

Calendar & Schedule Management

Own the calendar end to end: scheduling, conflict resolution, protected focus time, and a color-coded system across meeting types. Every day opens with a Start-of-Day brief.

Artifact: Start-of-Day brief
📧

Inbox Triage & Communications

Triage a high-volume inbox in real time with a label and classification system: act, draft, delegate, or surface urgently. Draft follow-ups and intros in the executive's voice.

Artifact: Triage framework
📋

Meeting Prep & Briefs

Build pre-meeting briefs for investor and sales calls so the executive walks in fully prepped: who, history, objective, and talking points.

Artifact: Pre-meeting brief
🤝

Relationships & Investor Database

Maintain a relationship database: contact records, meeting outcomes, follow-up cadences, and relationship stages, so no thread goes cold.

Artifact: Investor CRM schema
✈️

Travel & Event Coordination

Manage travel end to end (flights, hotels, ground, itineraries, day-of adjustments) and run events from guest lists and RSVPs through to post-event follow-up.

Artifact: Itinerary template
⚡

Extreme Ownership

I treat the role like it is my company. The founder's day runs smoothly, follow-ups never slip, and arrangements are handled with minimal intervention.

Operating principle
Automation Engineering

Not ChatGPT for emails. Real systems that run unattended.

Each build shown as Problem, Approach, and Result, with the architecture behind it. Client work is sanitized: the system and the outcome, never a client's data.

01

Invoice Processing Automation

60% less manual work
Problem
A finance operation handled invoices by hand. Slow, error-prone, hours lost weekly.
Approach
Mapped the workflow, ranked by time cost, and automated it end to end rather than in pieces.
Result
60% less manual processing, ~40% faster turnaround, 95% extraction accuracy.
Invoice inemail / PDF
→
OCR Extractline items
→
GPT Validateagent check
→
Exceptionhuman review
→
Store + Logcloud + ledger
02

AI Agent Orchestration Framework

Runs unattended
Problem
Single-prompt automations are brittle and break in production.
Approach
Separate the AI's reasoning from deterministic execution so the system stays predictable.
Result
Reliable multi-step automation, the backbone under every other workflow.
Requesttask / trigger
→
Reasoning LayerLLM agents
→
Review Gateguardrails
→
Execution LayerPython tools
→
Result + Learnself-improve
03

RAG Knowledge Base

85-92% answer accuracy
Problem
Information scattered across Drive, Gmail, Slack, and PDFs. Answers slow to find.
Approach
Centralize into a retrieval-augmented knowledge base with automated ingestion.
Result
85-92% answer accuracy on internal FAQs, 50-70% less search time.
SourcesDrive / Gmail / Slack
→
n8n Ingest + OCRautomated
→
Chunk + Embedvector index
→
Queryretrieve
→
Answercited
04

Executive Workflow Automations

Hours back each week
Problem
Recurring EA tasks eat hours: email classification, meeting notes, CRM updates.
Approach
Build a targeted agent for each task and chain them into the daily workflow.
Result
Several hours back per week, and nothing slips through the cracks.
Triggersinbox / meeting / CRM
→
Classify + Extractagents
→
Action Items+ drafts
→
CRM Update+ follow-ups
05

LinkedIn Content & Engagement Automation

Runs daily, hands-off
Problem
Staying visible on LinkedIn means daily content and genuine engagement. Hours of manual work, easy to let slip.
Approach
A multi-agent pipeline that researches, writes, self-reviews against quality gates, publishes, and engages, fully scheduled.
Result
Daily posts and quality-scored engagement (capped at 20/day) that run unattended, weekends auto-skipped.
Gathertopics + research
→
Createposts + AI visuals
→
Review + GateLLM scoring
→
Publishscheduled
→
Engage + Reportcapped, tracked
06

CRM & Lead Pipeline Automation

No lead left behind
Problem
A client's inbound leads arrived across forms, calls, and email, and slipped through manual follow-up. Slow response cost deals.
Approach
Automate intake, qualification, and follow-up inside the CRM so every lead gets a fast, consistent response.
Result
Faster first response, a follow-up cadence that never lapses, and fewer dropped opportunities.
Lead informs / calls / email
→
Classify + QualifyAI scoring
→
CRM Pipelineauto-stage
→
Draft Follow-upAI, on-brand
→
Team Handoff+ reminders
07

Recruitment Screening Automation

Hours to minutes
Problem
A client's roles drew high applicant volume. Screening CVs by hand was slow and inconsistent.
Approach
Score every CV against the job description on one rubric and rank candidates automatically.
Result
Shortlists in minutes instead of hours, consistent scoring, and faster time to interview.
Applications inCVs / forms
→
Parse + ExtractAI, structured
→
Score vs JDrubric
→
Ranked Shortlisttop matches
→
Recruiter Reviewfinal call
Toolkit

Stack

AI & Automation

n8nZapierClaude GPTGeminiPython OCRRAG / Vector DBCustom GPTs

EA & Operations

Google WorkspaceMicrosoft 365Notion AsanaCalendlySlack

CRM & Business

HubSpotGoHighLevelOdoo StripeFloat

Let's talk

Open to AI-powered Executive Assistant and operations roles with founders who want to move fast.

brypalma.biz@gmail.com LinkedIn Resume Full Portfolio (Notion)