Singa’s 23-person go-to-market team can now respond to inbound leads within minutes. Their reps open HubSpot and find leads already enriched, scored, and routed to the right person. The 45-minute morning sorting ritual is gone.
Singa provides karaoke software to restaurants, bars, and hotels worldwide. They handle hundreds of inbound leads per week across the US, Nordics, and UK. As paid advertising scaled, their inbound process couldn’t keep up.
“We needed automation that removes unnecessary manual work and improves response time to leads,” says Risto Mäkeläinen, Chief Commercial Officer. “But the automation must not weaken conversion or quality.”
Lead routing by gut feel doesn’t scale
Ly Hoang, B2B Growth Marketing Lead, started every day the same way. She spent 45 minutes manually sorting leads, figuring out which were worth pursuing, and deciding who should follow up. Leads lacked location data and finding the right information took time. High-value prospects got mixed in with low-priority ones.
The best leads go cold fast. Manual qualification couldn’t keep up with the volume.
Live in three weeks, built to keep improving
OstraconAI spent the first two weeks mapping Singa’s entire inbound process and setting up the technical foundation. That meant n8n for automation, Gemini for AI reasoning, and a full HubSpot integration. They identified 30 to 40 automation opportunities and ranked them by impact.
Week three, the system went live.
Every inbound lead now passes through an AI pipeline. It enriches contact data using external sources, qualifies each lead against Singa’s ideal customer profile and scores them by priority. The highest-priority leads trigger an instant Slack alert.
The whole process takes under two minutes. Every lead arrives in HubSpot with AI reasoning notes explaining the qualification and routing decision.
OstraconAI built feedback loops in from day one. A dedicated Slack channel lets the team flag errors. Prompts get refined weekly based on real data.
Reps now call the best leads the same day, with better data than before
Ly’s 45 minutes of manual sorting became single-digit minutes. Lead enrichment succeeds for over 80% of contacts. Reps open HubSpot and find leads scored, qualified, and routed. The highest-priority prospects get called the same day they come in.
Singa added an outbound engine to run alongside inbound
With inbound running and improving week by week, Singa added a second project: AI outbound. OstraconAI started building it in parallel, and the program went live in April 2026 across the US and UK markets.
The outbound system runs on two tracks. The first turns intent into action. When an ICP-fit company visits Singa’s website, the system enriches the contact, drafts a tailored first email with Claude, and routes it to Slack for one-click approval before sending via Instantly. Replies flow back into HubSpot automatically, sitting next to inbound leads in the same CRM.
The second track runs at volume. OstraconAI built segmented lead lists across US and UK karaoke bars, family entertainment centers, and bars with karaoke programming. The first wave covered roughly 10 000 prospects, each segment with its own messaging angle refined from Singa’s sales templates and call recordings.
The same principles apply: ship fast, keep a human in the loop where it matters, and tune weekly from real reply data. Inbound and outbound now feed the same HubSpot, the same reps, and the same feedback loops.
In the first two weeks alone, the US program reached around 2 000 prospects and produced 15 warm leads. The team expects that number to climb as it A/B tests subject lines, opening hooks, and CTAs against real reply data.
One AI GTM motion, two engines running in parallel
Three weeks from kickoff to live inbound system. A second engine built alongside it within months. That’s what happens when you start with the right problem.
Want to see what AI automation could do for your GTM process? Book a free strategy call with OstraconAI.
