- saas.wtf
- Posts
- Behind the Scenes: How Salesforge Is Using Smart Automation
Behind the Scenes: How Salesforge Is Using Smart Automation
By Karlis Kivlenieks, n8n Growth Hacker at Salesforge
Every founder is chasing an edge - ways to do more with less, grow faster, and serve customers better. It’s not a new ambition, but with AI now in the spotlight, the push for smarter, leaner operations is stronger than ever.
At Salesforge, where we’re building cutting-edge sales infrastructure, and we are using automation to help our team stay lean and move fast.
We rely heavily on n8n. In just the last 7 days, we’ve executed over 15,000 automation runs, and in the past 4 months alone, we’ve built 200+ workflows - and counting.
From day one, our CEO, Frank Sondors, made it clear: automation is not just a nice-to-have - it’s part of our company's DNA. He encouraged the team to share any ideas they had for automation directly in Slack.
That’s how we created our internal “Automation Idea Vault” - an open space where anyone at Salesforge can submit ideas. Each week, we review and prioritize the ones with the biggest potential impact on our day-to-day operations and long-term growth.
Most of our automations support Sales, Customer Success, and Operations, giving our small team superpowers and freeing them up to focus on what really matters and at what they are great at.
A Few Things to Remember When Building Automations
If you’re thinking about introducing automations in your company - or doing more of them - here are a few suggestions:
1. Make Someone Responsible for Building Automation
I’m a big believer in the “citizen development” movement. But, after teaching hundreds of people no-code and low-code tools, reality is that just because anyone can build automations doesn’t mean they will. People default to what they’re good at - and passionate about. So, instead of asking your sales team to build internal tools, let them focus on selling. Ideally, assign a dedicated person to handle automation work.
2. Tools Like n8n Are Just the Engine
n8n (or any automation tool) connects your apps and data - but it’s not the whole machine. You’ll still need:
A database to store data from automations (e.g., statuses, timestamps, metadata).
Other services from where you will get information, API providers, etc.
Think strategically about which services are worth paying for - and skip the nice-to-haves. Don’t fall for every shiny tool.
3. Don’t Use Google Sheets as Your Automation Backend
Google Sheets is great for manual tracking, but it’s too slow and fragile for serious automation.
If you’re comfortable with a more technical setup, go with Supabase. Otherwise, Airtable is a solid choice - but beware of high costs if you’re working with large volumes of data.
Thanks to the algorithm, my LinkedIn is flooded with screenshots of beautiful automations built with n8n. But building something that looks good in a demo is very different from building something that scales reliably.
Automations take time to build, test, and refine. Don’t be misled by polished proof-of-concepts.
5. AI Isn’t Always Necessary
AI agents are trendy, but not every workflow needs AI. If your process is deterministic (clear rules, no judgment calls), a classic automation setup will be more efficient, faster, and cheaper.
6. Plan for Scale - Early
We built quickly and ended up hitting n8n cloud’s subscription limits fast. If you think scale is even a possibility, go self-hosted from the start.
We now run our own n8n instance on Railway - and if I could go back, I’d have made that move much earlier.
From AI Hype to Real Use: How We Built Our Own Voice Agent to Support Customer Success
We’re long past the days when Sundar Pichai’s AI made a dinner reservation on stage. At Salesforge, we’ve built our own voice agent - not as a flashy demo, but as a real automation that drives value daily.
Despite serving thousands of customers, our Customer Success team is just two people (plus support agents). We help them by having layered multiple automations across the customer journey - especially to help CS intervene at the right moments.
We’ve built automations that:
Track customer growth (e.g., subscription increases),
Monitor payment issues or churn signals,
Alert CS when action is needed.
This enables our team to reach out at the right time, with context and purpose.
Post-Demo Outreach Automation
Previously, if there was no communication after a demo, we’d trigger follow-up emails. But recently, we wanted to take it further.
Now, with the help of n8n, we’ve built an AI voice agent that reaches out to demo prospects with a personalized, human-sounding call.
And yes - it speaks in the voice of our CEO. Say hello to: Voice Agent Frank (not to be confused with the Agent Frank product we offer clients).
How It Works:
Preparation (n8n-powered):
After each demo call, we schedule a follow-up window.
Before any call happens, we check:
Has the prospect communicated with our CS or CEO?
Are they already a customer?
Is this the right timing based on the last interaction?
Email access (read-only) + notes from the demo are used to assess context.
Decision Layer:
Our AI agent reviews all the info.
If a call isn’t needed, it updates CS with status and suggests next steps.
If a call is needed, we trigger the voice call via Retell.
Voice Agent Behavior:
The agent knows who it’s calling, what was discussed during the previous call, and which tool the prospect was interested in.
It’s not a sales call - the agent’s job is to help, not close.
It answers questions, offers support, and suggests next steps.
Scheduling the Follow-Up Call (With More Automation)
If the prospect wants to speak with CS again:
The agent asks for preferred times (aware of their timezone).
This time is sent to n8n, which checks availability on the CS calendar.
If unavailable, another AI proposes close alternatives.
Once agreed, the agent confirms participants and:
Triggers an n8n workflow to schedule the meeting,
Sends invites to both sides.
Here is a real example:
Automation Is About More Than Just Saving Time
At Salesforge, automation isn’t just about saving time - it’s about amplifying impact. By combining smart workflows with thoughtful human oversight, we’re building a lean, fast-moving engine that scales without sacrificing quality. Whether it’s AI-powered voice agents or back-end automations that never see the spotlight, every piece plays a role in helping us grow without bloating the team.
The takeaway? Automation works best when it’s practical, intentional, and designed to support - not replace - the humans behind the mission.