Most founders doing outbound are not short on effort. They are short on time. Research before every message, writing personalised copy, tracking who needs a follow-up, logging to the CRM. None of it requires creative thinking. All of it takes hours that should be spent in conversations.
toflow.ai is built around a different premise: connect it to Claude or ChatGPT once via MCP, and then run your entire outreach operation by typing what you want. The prospecting, enrichment, sequence building, messaging across email, LinkedIn, and WhatsApp. All of it runs through prompts. You stay focused on the conversations. Claude handles everything before and after.
Claude as your outbound function
The one-time setup is connecting your LinkedIn and email accounts in Settings, Accounts, then linking toflow to Claude or ChatGPT via MCP. No credentials needed. Just add the toflow MCP URL and authorise your workspace. Full setup is on the Claude and ChatGPT pages. After that, every action runs through a prompt.
What that looks like in practice:
"Find 40 Head of Sales at B2B SaaS companies in India with 50 to 200 employees that are actively hiring SDRs. Check which ones I'm already connected to on LinkedIn, remove duplicates from my existing lists, and enrich the rest with verified emails."
Claude runs the LinkedIn search through toflow, filters by connection status, cross-checks existing contacts, enriches the new ones across 8+ data sources, and returns a clean list, all in one conversation. What a researcher would spend a day on runs in minutes.
You can go further in the same session:
"Read what each of these contacts has posted on LinkedIn in the last 30 days and note any that have written about outbound sales challenges."
Claude reads their activity through toflow and flags the ones with relevant signals. Those contacts get a connection request that references their post specifically, not a template with a first name dropped in. The personalisation is real because it draws from what each person actually wrote.
Review before sending: speed with control
When you enroll contacts, toflow generates personalised message content per prospect at enrollment, pulling from their LinkedIn profile, recent posts, company stage, and the sequence prompts. You can choose to enroll with needs-approval mode, which queues the enrollment for review before the sequence runs, or enroll directly if you want it to start immediately.
For a founder, the review step matters. Your name is on every message. Enrolling with needs-approval means you stay in control of what goes out without writing from scratch. You scan the generated content per enrollment, approve the ones that land well, regenerate the ones that do not, and the sequence runs.
"Regenerate the connection request for the three contacts where the angle is too generic. Reference their company's recent funding instead."
Claude rewrites those three. You approve. Done. The whole review takes ten minutes across fifty contacts.
What a week of founder outreach actually looks like
Monday. Open Claude and describe this week's target: a specific ICP segment, a new geography, a list from a recent event. Claude searches, enriches, and builds the list inside toflow. You spend fifteen minutes reviewing it, removing anyone who does not fit, and the list is ready.
Tuesday. You enrolled with needs-approval mode, so the enrollments are queued for review. You scan the generated content, approve in batches, regenerate a few where the angle does not land. LinkedIn connection requests start going out during business hours in each prospect's timezone, spaced to stay within daily limits.
Wednesday. Replies come into the unified inbox, where email, LinkedIn, and WhatsApp sit together. You ask Claude to categorise them and draft responses for the interested ones. You review, send, and update deal stages in the CRM with a single prompt.
Thursday. A LinkedIn post from a thought leader in your space got 200 comments. You paste the URL into Claude: "Pull everyone who commented on this, filter for VPs at SaaS companies with 50 to 500 employees, enrich their emails, and add them to the India list." Claude returns 18 qualified contacts ready for the next enrollment run.
Friday. "How is this week's sequence performing? Which step has the lowest reply rate?" Claude pulls analytics from toflow and gives you a breakdown. You adjust the copy on the underperforming step, ask Claude to regenerate those drafts, and queue next week's run.
That is a full outbound function running five days a week. No SDR hire. No tool stack. One conversation with Claude.
Who this works best for
- Pre-PMF founders running outreach to validate ICP and book discovery calls before making a sales hire
- Post-seed founders who have early pipeline but need to systematise before bringing on an SDR
- Technical founders who want AI-driven outreach without managing a stack of separate tools
- Founders selling into India, Southeast Asia, LATAM, or the Middle East where WhatsApp or LinkedIn is the primary channel alongside email
Is toflow.ai the right fit?
Makes sense if:
- You want to run outreach by describing who you want to reach rather than navigating multiple tools manually
- You need email, LinkedIn, and WhatsApp from one place so nothing falls through when you are stretched across product and fundraising
- You want AI to handle prospecting, enrichment, and execution while you focus on conversations that matter
Probably not the right fit if:
- You are pre-product and not yet doing any outbound. Validate the idea in direct conversations first
- You only need basic email sequences with no LinkedIn or WhatsApp component
Book a demo now. Two weeks free trial, no credit card required.