Relevance Live 07: Building Tools with Multiple Data Sources (Step-by-Step Agent Tutorial)

Oct 14, 2025Channel
AI Analysis
Data from YouTube Data API v3Updated Just now
Relevance AI
Relevance AI

10.9K subscribers

View Channel

Video Overview

Video Details

Published8 months ago
Duration54:07
Video IDbP7SMGr72ZM
Languageen
CategoryHowto & Style
PrivacyPublic
Made for KidsNo
Video TypeRegular Video

Performance Metrics

Views56
Likes3
Comments0
Engagement Rate5.36%
Likes per 100 views5.36
Comments per 1K views0.00

Description

In this session of Relevance Live, Senior Solutions Engineer Brian walks through how to build tools that connect multiple data sources - from Salesforce to Outreach - and automate the “last mile” of your agent workflows. Follow along as Brian builds an Outbound Email Tool from scratch inside Relevance AI, showing: - How to chain API calls - Extract and transform data dynamically - Use LLM steps to handle edge cases - And troubleshoot real-time debugging Perfect for anyone exploring how to operationalize AI agents or integrate them with enterprise tools. 🕒 Timestamps 00:00 – Intro: Welcome from Chelsea & Brian 01:40 – Overview: Why “the last 10%” matters in agent building 03:00 – Use Case Setup: BDR Agent with Salesforce + LinkedIn + Google 06:10 – Generating outbound email sequences automatically 09:00 – Taking it from manual to scalable: integrating with Outreach.io 11:40 – Building the Outreach Tool from scratch 13:30 – Defining inputs and connecting Salesforce credentials 16:00 – Setting up Outreach authentication 18:10 – Retrieving the lead’s email address via Salesforce API 22:00 – Using LLM steps to extract clean outputs 25:30 – Creating dynamic API calls with variables 30:00 – Debugging the Outreach prospect search flow 33:10 – Structuring JSON payloads for email sequences 37:00 – Patching email sequences to Outreach prospects 41:00 – Troubleshooting errors live (and why it’s okay) 44:20 – Audience Q&A: LinkedIn search access, Relevance Chat 48:00 – Behind the scenes: how chat.relevance.ai helps with API discovery 52:00 – Discussion: Tools vs. Agents – when to use each 55:00 – Final Tips: Thinking in inputs/outputs, chaining steps 57:00 – Outro & upcoming Relevance Live sessions 🧠 Key Takeaways 1. Think of tools as modular blocks that handle specific automations. 2. Leverage Relevance Chat to design or troubleshoot APIs faster. 3. The “last mile” of automation (integration & action) is where most teams win.

Related Videos

More videos from Relevance AI