Nextvant · Aug 2026 – Present
AI agents that find buyers before they raise their hand
Nextvant needs a steady flow of sales conversations with companies that could use its GTM services. I built a chain of AI agents that spots companies showing signs they need help, checks they're a good fit, writes a personal first message and sends it by email and LinkedIn. So far it has scanned 3,000+ accounts and booked 20 meetings.
- Company
- NextvantGTM & HubSpot services for B2B businesses
- My role
- Founder's Office Associate
- When
- Aug 2026 – Present
- Tools
- Apify · Clay · Apollo · Slack · OpenAI/Claude APIs · Instantly · HeyReach · n8n · Make · HubSpot
- SourceNextvant, Nov 2026
- SourceNextvant buying-signal engine
The problem
A services business lives on conversations with the right companies at the right time. Doing that by hand is slow: you have to research each company, decide whether it's a fit, find the right person and write a message that doesn't read like a template. Generic messages get ignored, and careful ones don't scale.
The goal was a system that does the research and first draft automatically, so outreach can be both personal and frequent.
How the engine works
Each step hands its output to the next one:
01
Scrape buying signals
Apify
02
Enrich and qualify accounts
ClayApollo
03
Route qualified accounts to Slack
Slack
04
Generate personalised messages
OpenAI/Claude APIs
05
Send across email and LinkedIn
InstantlyHeyReach
- Spot the signal. Apify collects buying signals: public signs that a company may need GTM help now.
- Check the fit. Clay enriches each account with company and contact details and scores it, and Apollo fills in contact information. Accounts that don't fit stop here.
- Route it. Qualified accounts are posted to Slack, so the team sees every lead the engine finds.
- Write the first message. OpenAI and Claude models draft a message specific to that company and its signal, not a mail-merge template.
- Send it. Instantly sends the email sequence and HeyReach handles LinkedIn outreach.
What I did
- Designed the flow from signal to sent message, deciding which tool does which job.
- Built the agents and the connections between Apify, Clay, Apollo, Slack, the AI models, Instantly and HeyReach.
- Kept it running alongside client work, as part of my founder's office role.
This is one of several automations I've built at Nextvant. I also build workflows in n8n and Make, and intent-data workflows in HubSpot.
What I'd do next
Track which signals actually lead to meetings and weight the scoring towards them, so the engine gets sharper with every send.
Key terms (5): new to the jargon?
- Buying signal
- A public sign that a company may need what you sell right now. Reaching out at that moment makes a message far more relevant than a cold pitch.
- Enrichment
- Filling in the details about a company or person, such as size, industry, role and contact details, so you can judge fit and reach the right person.
- Qualified account
- A company that matches the profile of a good customer and is worth contacting.
- Outbound
- Sales outreach you start yourself (emails, LinkedIn messages), as opposed to inbound, where the customer finds you first.
- Intent data
- Data that suggests a company is actively researching a problem or product, used to decide who to contact and when.
What's next
