The back office for the agents the big carriers forgot.
Somewhere in New York, on a Tuesday in October, an independent Medicare agent opens a laptop. Last year this was the cruelest month: a desk buried under carrier portals, commission spreadsheets that disagreed with each other, and a stack of clients quietly drifting toward a competitor's plan. This year the screen shows one thing - a single book of business, every client status visible, the at-risk ones already flagged. The agent has not called IT. There is no IT.
That calm is the entire product. Spark Advisors is a technology-enabled Medicare brokerage - in the trade, a national marketing organization - that decided the independent agent deserved the kind of software usually reserved for people who work at the carrier, not against it. Today more than 10,000 brokers run on it. In 2025 those brokers pushed roughly 250,000 enrollments through the platform.
More than 60 million Americans are on Medicare, and a great many of them buy their plans through an independent agent - a one-person business, or a small agency, juggling dozens of carriers. The carriers built slick apps for themselves. The agent got a fax machine and a prayer.
The result was an industry running on duct tape: contracting paperwork that took weeks, commission statements that arrived wrong, "ready to sell" status that nobody could see, and clients who churned without anyone noticing until the check stopped coming. It was, in the great tradition of important and boring work, ignored by almost everyone with the skills to fix it.
In 2020, James Jiang, Alex Zhang, and Byron Edwards started Spark on a premise that sounded almost contrarian: the independent Medicare agent is not going away, and the tools that serve that agent are worth building well. Jiang, who had worked in finance since 2014 and co-founded an earlier startup, took the CEO seat. Zhang runs product as CPO. Edwards runs operations as COO.
The legal entity is Spark Health, Inc.; the brand everyone says out loud is Spark Advisors. The distinction matters less than the posture - this is a healthcare company that behaves like a software company, in a market most software people had written off.
Spark's pitch is that it is the only such organization with an all-in-one platform serving brokers, agency principals, and call centers at once. In practice that means the unglamorous work - contracting, licensing, reconciliation, retention - now happens in software instead of in someone's inbox.
Contracting, finance, ready-to-sell tracking, onboarding visibility, performance reporting, and commission reconciliation in one place.
Lead capture, automated retention campaigns, plan comparison tools, and AI Churn Defense that flags clients before they leave.
Lead queries, production reconciliation, and performance analytics for telephonic and virtual sales teams.
Recruitment, marketing support, CMS-compliant assets, and quality goal planning handled on the agent's behalf.
The case for Spark is not a vision deck; it is a retention rate. Agencies that move onto the platform report roughly 40% growth in enrollments. The reported client retention rate sits at 84%, and the Net Promoter Score - in insurance software, a category not famous for being loved - is 91.
The money came from investors who know the space. American Family Ventures - the venture arm of an insurer - has backed Spark across rounds, alongside Primary Venture Partners from the very beginning and Viewpoint leading the Series B. That is the kind of cap table that suggests the boring market is, in fact, a big one.
Spark's stated goal is to transform how people navigate healthcare by giving independent agents the technology and service to do it well. The interesting part is what it refuses to remove: the agent. Plenty of insurtech tried to cut the human broker out entirely. Spark bet the opposite - that the relationship between an agent and a 72-year-old choosing a plan is the valuable thing, and the paperwork around it is the part worth killing.
So the company sells to the people the carriers treat as an afterthought, and treats them like the main event. The platform handles contracting and reconciliation and churn signals; the agent handles the conversation. It is a tidy division of labor, and it is doing roughly a quarter-million enrollments a year.
The math is unsentimental: the population is getting older, Medicare keeps growing, and plan choices multiply every year. Someone has to sit with people and translate. For 2026, Spark is building tools to serve clients by phone and video, year-round rather than just at enrollment, with AI surfacing opportunities across an agent's entire book. The bet on the independent agent is now a bet that the agent, well-equipped, scales.
Which brings us back to that Tuesday in October. The agent closes the laptop early - the at-risk clients have been called, the commissions reconciled themselves, the contracting went through last week. The cruelest month turned out to be just a month. That is the whole point of Spark Advisors: not to replace the person who picks up the phone, but to make sure that when they do, nothing behind them is on fire.
Figures (enrollments, retention, NPS, funding, headcount) are drawn from Spark's own announcements and public press; revenue and some round details are third-party estimates and should be treated as approximate.