Skip to main content
The Sixtyfour Agent is an AI assistant built into the platform that translates natural language queries into fully functional workflows. Instead of manually configuring blocks and connections, describe what you need and the Agent builds it — search, enrichment, filtering, output, and everything in between. The Agent has access to the entire Sixtyfour platform including: all workflow blocks, the Search API, enrichment endpoints, notebooks, integrations, and your account settings.

Getting Started

Open the Agent from the sidebar in the Sixtyfour dashboard. Enter a query describing what you want to accomplish, and the Agent handles the rest — creating workflows, running searches, and returning results. You can then chat with the Agent to refine your results. Ask follow-up questions, adjust filters, add enrichment steps, or change the output format until the pipeline matches your needs.

What the Agent Can Do

Search for People and Companies

The Agent searches across Sixtyfour’s dataset using natural language. Find leads by job title, company, location, seniority, or any combination — or search for companies by industry, size, funding stage, and other criteria. Example queries:
  • "Find CTOs at Series B AI startups in San Francisco"
  • "Search for marketing directors at companies with 50–200 employees"
  • "Find competitors to Snowflake in the data infrastructure space"

Build and Manage Workflows

The Agent creates end-to-end workflows from natural language. It selects the right blocks, configures them, connects them in the correct order, and handles data type conversions automatically. The Agent can also read, edit, run, and cancel existing workflows. It inspects a workflow’s current state before making changes to prevent conflicts. Example queries:
  • "Build a workflow that finds VPs of Sales at fintech companies, gets their emails, and saves verified results to a notebook"
  • "Add a filter to my workflow that removes leads without a LinkedIn URL"
  • "Run my lead enrichment workflow and notify me in Slack when it finishes"

Construct Intelligence Pipelines

The Agent assembles multi-step pipelines using blocks across every category:
CategoryExamples
DiscoveryPeople search, company search, GitHub contributors, Google Maps businesses, job postings, Y Combinator companies
EnrichmentAI-powered people and company research, email finding, phone finding, email verification, reverse lookups
ExtractionWeb scraping, subdomain enumeration, backlink analysis
TransformFiltering, deduplication, text generation, data type conversion, grouping
AILLM-powered classification, extraction, or structured output per row
OutputSave to notebooks, send Slack messages, fire webhooks

Handle CSV Data

Upload a CSV file and the Agent incorporates it into a workflow automatically. The Agent generates inline data on the fly and provides real-time feedback directly in chat.

Manage Notebooks

The Agent lists, inspects, and creates notebooks — Sixtyfour’s managed spreadsheets for storing enriched data. Workflows output directly to notebooks, and the Agent can set that up as part of any pipeline.

Optimize for Cost

The Agent proactively suggests ways to reduce credit usage. It places filters before expensive enrichment steps, warns before running high-cost operations, and requests only the fields you need.

Access Account Information

The Agent shows your credit balance, usage statistics, team members, and organization details. It can generate invite links for new members.

Review Run History

The Agent retrieves execution logs for any workflow, showing run status, timestamps, rows processed, and credit costs.

Research the Web

The Agent searches the web for up-to-date information when needed — identifying companies in a specific market, confirming current data, or informing workflow design with real-time context.

How the Agent Works

The Agent is built around a few core principles:
  • Type-aware. Sixtyfour workflows operate on two data types — Lead (people) and Company (organizations). The Agent ensures blocks connect in valid sequences and inserts conversion blocks (like Extract Leads) automatically when a type change is required.
  • Safety first. The Agent reads a workflow’s current state before making edits, preventing conflicts with in-progress runs or existing block configurations.
  • Proactive. The Agent suggests optimizations and flags potential issues — like placing filters before expensive enrichment steps — rather than executing blindly.
  • Honest. If the Agent lacks enough information to answer a question or build a workflow, it says so rather than guessing.

Example Use Cases

GoalExample query
Build a prospecting list"Find 50 engineering managers at healthcare companies in Boston and get their emails"
Research a market"Search for AI infrastructure companies founded after 2020 with more than 50 employees"
Enrich existing data"Take my leads notebook, enrich each person, filter for verified emails, and deduplicate"
Monitor competitors"Find all developers who starred Competitor's GitHub repo and get their LinkedIn profiles"
Automate outreach prep"Build a workflow that finds decision-makers at target accounts, researches each person, generates a personalized intro, and saves everything to a notebook"

Agent Tips

  • Be specific. The more detail you give — job titles, locations, company criteria, desired output — the better the Agent’s first result.
  • Iterate. The Agent is conversational. Refine your query, ask it to adjust filters, or add steps to an existing workflow.
  • Let the Agent optimize. It knows which block ordering minimizes credit usage. Describe the outcome you want and let it choose the pipeline structure.