Event Organizer
Event Speaker Outreach — From Luma List to Confirmed Speakers in One Week
Key Takeaway
An AI agent researches speakers, sends personalized multi-channel outreach, and tracks the pipeline in Google Sheets. Secured 3 confirmed speakers for our Barcelona meetup in one week.
The Problem
We were organizing the Mr.Chief Barcelona meetup. Venue booked. Date set. Luma page live. One problem: no speakers.
Finding and securing speakers for a niche tech event is a multi-step nightmare:
- Research: Who's spoken at similar events? Who's publishing in the space? Who's local?
- Contact info: LinkedIn profile, email, Twitter — find at least two channels
- Personalization: Generic "would you like to speak?" emails get ignored. You need to reference their work specifically
- Multi-channel: Some people respond on LinkedIn. Others email. You need to hit both
- Follow-up: The first message rarely gets a reply. You need a 3-touch sequence
- Tracking: Who replied? Who needs a follow-up? Who confirmed? When?
For 10 target speakers, that's easily 15-20 hours of work. We had a week. And I had other things to do.
The Solution
The agent took a list of 10 target speaker profiles, researched each one thoroughly (recent talks, publications, social presence), generated personalized multi-channel outreach sequences (LinkedIn + email), executed the sequences with proper timing, and tracked the entire pipeline in a Google Sheet. One week. Three confirmed speakers. Zero hours of my time on outreach.
The Process
Step 1: Define the target list
markdownShow code
# Speaker targets — Mr.Chief Barcelona Meetup
Event: Mr.Chief Barcelona Community Meetup
Date: April 15, 2026
Venue: MOB Coworking, Bailèn 11
Format: 3 x 20-min talks + networking
Topic: AI Agents in Production — Real Architectures
Targets:
1. Elena Martínez — AI Lead @ Glovo, Barcelona. Spoke at PyBCN about LLM orchestration.
2. Carlos Ruiz — CTO @ Factorial. Building internal AI tools at scale.
3. Ana Soler — Founder @ AgentOps.ai. Agent monitoring platform.
4. Jordi Puig — Sr. Eng @ Typeform. Published on multi-agent patterns.
5. Marc Dubois — VP Eng @ Datadog (Barcelona office). Observability for AI.
...
Step 2: Deep research per speaker
bashShow code
# Agent researches each target:
# - Recent conference talks (YouTube, Luma, Meetup.com)
# - Published articles/blog posts (personal blog, Medium, company blog)
# - LinkedIn posts in last 90 days
# - Twitter/X activity on relevant topics
# - Mutual connections in our network
# Output per speaker:
## Elena Martínez — Research Brief
- Last talk: "Orchestrating LLMs at Glovo" (PyBCN, Jan 2026)
- Recent LinkedIn post: scaling agent reliability (2.3K likes)
- Blog: "Why We Stopped Fine-Tuning" (company blog, Feb 2026)
- Mutual connections: 2 (Pierre, Yann)
- Email: found via company domain pattern
- Angle: She's talked about orchestration — pitch the "real architectures" angle
Step 3: Personalized outreach sequences
markdownShow code
# Outreach template — customized per speaker
## LinkedIn Connection Request (Day 1)
"Hi Elena — I caught your PyBCN talk on LLM orchestration at Glovo.
We're running a similar meetup in Barcelona on April 15 focused on
AI agents in production. Your perspective on orchestration at scale
would be exactly what the audience needs. Mind if I share details?"
## Email (Day 2)
Subject: Speaking invite — AI Agents in Production (Barcelona, Apr 15)
Hi Elena,
I'm Bilal, founder of Mr.Chief (open-source AI agent framework).
We're hosting a community meetup in Barcelona on April 15 at MOB
Coworking — "AI Agents in Production: Real Architectures."
Three 20-minute talks plus networking.
Your PyBCN talk on LLM orchestration was one of the most practical
takes I've seen. We'd love for you to share what you've learned
scaling agents at Glovo — particularly the reliability patterns
you mentioned in your recent LinkedIn post.
Format: 20-min talk, ~50 attendees, no recording unless you want it.
Luma page: [link]
Would this work for your schedule?
Best,
Bilal
## Follow-up (Day 5, if no response)
Subject: Re: Speaking invite — AI Agents in Production
Hi Elena — just following up on the Barcelona meetup invite (April 15).
Happy to jump on a 5-min call if you have questions about the format.
No pressure either way.
Step 4: Pipeline tracking in Google Sheets
bashShow code
# Agent creates and maintains the tracking sheet
gog sheets create "Barcelona Meetup — Speaker Pipeline"
# Columns: Name | Company | LinkedIn Sent | Email Sent | Follow-up |
# Response | Status | Confirmed Date | Talk Title
gog sheets update "Barcelona Meetup — Speaker Pipeline" \
--range "A2:I2" \
--values "Elena Martínez,Glovo,Mar 5,Mar 6,,,,,"
# Updated as responses come in:
gog sheets update "Barcelona Meetup — Speaker Pipeline" \
--range "F2:I2" \
--values "Mar 7 — interested,Confirmed,Mar 8,Orchestrating Agents at Scale"
Step 5: Response handling
When a speaker replies (via email or LinkedIn), the agent:
- Detects the response and categorizes it (interested, declined, needs info)
- Updates the tracking sheet
- Notifies me on Telegram: "Elena replied — she's interested, wants to know about recording policy"
- Drafts a reply for my approval
- Coordinates logistics (title, bio, AV needs) once confirmed
The Results
| Metric | Manual Outreach | Agent Outreach | Change |
|---|---|---|---|
| Time spent on outreach | 15-20 hours | 45 min (reviews only) | -96% |
| Speakers contacted | 10 | 10 | Same |
| Response rate | ~30% (generic) | 70% (personalized) | +133% |
| Speakers confirmed | 1-2 (typical) | 3 | +50-200% |
| Time to first confirmation | 5-7 days | 3 days | -50% |
| Follow-up consistency | Inconsistent | 100% on schedule | Perfect |
| Pipeline visibility | Scattered notes | Live Google Sheet | Centralized |
Three speakers confirmed in one week. Elena from Glovo, Jordi from Typeform, and Ana from AgentOps.ai. The Barcelona meetup had 26 attendees — not huge, but exactly the right audience. The personalization made the difference. Every speaker who replied said something like "I can tell you actually read my stuff."
Try It Yourself
- Create your target speaker list with names, companies, and any known angles
- Let the agent research each target — recent talks, posts, publications
- Review the personalized outreach sequences before sending (important for first events)
- Set up the tracking sheet and let the agent maintain it
- Handle responses through the agent — it drafts replies, you approve
The key lesson: outreach personalization at scale is the agent's superpower. A human can deeply personalize 2-3 emails before fatigue sets in. The agent does 10 with the same depth, and it researches faster. The combination of research + personalization + consistent follow-up is what turns a 30% response rate into 70%.
Three speakers. One week. Zero hours of outreach work. The meetup basically organized itself.
Related case studies
Operations Lead
Populating Google Sheets From Agent Data — Live Dashboards Without Code
AI agents write operational data directly to Google Sheets via gog CLI, creating live dashboards that update themselves — zero manual data entry, zero code, zero BI stack.
BD Lead
LinkedIn Outreach at Scale — 50 Personalized Connections Per Week
From Luma event attendees to 50 personalized LinkedIn connections per week — zero active time, 48% acceptance rate, and 10-14 meetings booked per month.
Founder
Slack-to-Telegram Bridge — Team Messages Without Opening Another App
An AI agent bridges Slack and Telegram bidirectionally — forwarding only messages that matter to Telegram, posting replies back in Slack — zero Slack checks, 4-min response time.
Want results like these?
Start free with your own AI team. No credit card required.