Business Case: Byte Insurtech × VoiceB.ai
How a digital insurance distributor turned web-form leads into qualified, priced conversations — and prepared sales around the clock — with autonomous voice agents.
The company
Byte Insurtech is a digital insurance distributor in Spain. It markets several insurance products online and generates demand through web forms: a prospect lands on a product page, leaves their details, and expects a call.
The challenge
Online insurance demand is won or lost in the first minutes — the odds of ever qualifying a web lead fall sharply once the first few minutes pass (HBR / MIT, The Short Life of Online Sales Leads). A lead that waits is a lead that cools. And Byte’s demand has three properties that make a human-only phone team hard to scale:
- It arrives 24/7 and in spikes — including nights and weekends, when advisors are off.
- Every product qualifies differently — who is eligible, which age range, student vs worker, resident vs expat — all before a price can even be quoted.
- Several products run at once, each needing its own script, price and objection handling.
With a human-only team the outcome is familiar: leads called late or not at all, advisors spending their hours qualifying instead of closing, and nothing happening out of hours.
The solution
We deployed autonomous sales voice agents (ASVAs) — one per product line — each triggered the instant a web lead arrives (click-to-call) and reachable on the same inbound number. Each agent is built in Agent Studio with the four-part structure we use on every campaign:
- Goal — qualify the lead against the product’s rules, present the price, and either prepare the sale end to end or route a qualified, priced lead to a Byte advisor.
- Script — a short, protected opening, then eligibility, price and data capture.
- Guardrails — eligibility checks are enforced before any quote (age range, student-not-worker, residency); no price is given outside the rules.
- Objections — price first, plus the real coverage questions (what’s included, waiting periods, repatriation).
When a lead comes in, the agent calls within seconds, confirms eligibility, presents the monthly and total price, handles the price and coverage objections, and captures the data needed to activate the policy. If the sale can be prepared end to end, it is. If the customer asks for a person, or a question needs a specialist, the agent hands over a lead that is already qualified and priced — and outside office hours it books the advisor callback and still captures everything. Byte handles the live call routing on its side.
The results
One campaign — Byte’s student health-insurance line — over a 30-day window (12 Oct – 11 Nov 2026):
Sales were prepared by the agents; final activation was completed by Byte or the customer. No sales were fully autonomously closed in this reporting window.
Close to half of the live conversations produced a commercial outcome — approximately 45% of live conversations: a sale the agent prepared, or a lead handed to an advisor already qualified with the price confirmed. None of it waited for office hours.
This fits a wider pattern across our book. In energy, our agents close 30.4% of valid inbound calls fully autonomously for Holaluz — the same engine, a different vertical.
Why it matters
The advantage is not that an agent is cheaper than a person. It is capacity. An agent calls every lead in seconds, 24/7, across every product at once, and never leaves the top of the funnel unattended. Advisors stop qualifying and start closing leads that are already priced and ready. And because each new product is just another agent in Agent Studio, adding a line is a configuration — not a hire.
Byte measures what matters — sales prepared and qualified leads delivered, not call activity — and the next step is to move more of its traffic to autonomous handling and raise the share the agents close end to end.
— Alex Bisbe, Co-founder & CEO, VoiceB.ai

