Benefits administration for insurance agencies.
An individual coverage HRA is a funding decision; so is a level-funded group plan. They’re all benefits administration problems, and they’ve been solvable for years. What makes them hard is running each one in a different system. Abra administers whatever a group chooses to offer, in one place, with your agency in front of it.
Figures reflect the platforms this team has designed and operated since 2015, and while the stack behind them has changed more than once, what we learned about benefits administration carried over each time.
It’s a fair question to put to anyone asking to administer your book, and in this category the answers fall into a narrow range. Most ICHRA platforms were incorporated after the 2020 regulations created the market, which means the whole company is younger than the arrangement it sells.
Built after 2020 for ICHRA specifically. Genuine focus, but no group book to learn from and no history with the messy cases that make up most of a renewal.
Administration expertise, wrapped in a platform built later to productize it. The service model usually shows through, and it tends to want a relationship with your client.
Modern tooling makes it possible to ship a convincing benefits platform in a year. What it can’t manufacture is the plan data, the carrier behavior, and the edge cases you only get by living through them.
Abra is the fourth answer. This team has been administering group benefits since 2015 and added ICHRA to a platform that was already running. The institutional knowledge predates the software by nine years.
The original enrollment platform, in production since May 2015 with real brokers, employers, and carriers. It’s where the 15,000 open enrollments and 450,000 qualifying events above come from.
An ICHRA-native platform launched alongside Beacon — a deliberate parallel-architecture test, not a replacement. Zero to 3,000+ employees offered ICHRA between 11/1/2025 and 1/1/2026.
Both streams converged here, and every structural decision traces back to a specific lesson from Beacon or V1. The stack is new. The requirements it satisfies are the ones we accumulated running the other two.
We use modern tooling to build quickly. No point pretending otherwise — the question is what it’s pointed at. Ours is pointed at a decade of our own enrollment history, our own plan and rate data, and our own record of how carriers behave when a file fails. It speeds up work we already knew how to do. Build the same way without that history and the software is inferring how benefits administration works, which is a poor substitute for having processed 450,000 qualifying events yourself.
Enrollment technology started with a modest goal: retire the paper application and give a group administrator some visibility into their own benefits, through the broker who sold them. Somewhere along the way the category decided its job was engagement — more dashboards, more modules, more reasons for the employer to log in and talk to the platform instead of a person. We think that was the wrong turn, and most of the complexity in this market exists to justify it.
Class-based eligibility, mid-year funding changes, ancillary and worksite lines, ACA rate math, carrier-specific file quirks — all of that belongs in the engine. What reaches your screen, or your client’s, should be a short list of decisions and a clear picture of where things stand.
Your client has a payroll system and probably an HRIS they chose for reasons that had nothing to do with benefits. We’re not trying to replace either one. Benefits administration is a specific job, and doing it well next to their existing stack beats asking them to abandon it.
Groups reach the platform through an agency, and they reach their answers through one too. We build the tools that make you faster and better informed. We don’t build the ones that route around you, because we don’t have anywhere to route them to.
Technology should administer the benefit. A person should explain it.
Whatever benefits look like in ten years, no group of any size will administer them without software. But choosing between two networks when a child needs a specific specialist is not a decision to hand to a chatbot, and it’s far too consequential to make without good information in front of you.
So we’re building toward the version where the platform carries the administration and the education, and the broker carries the conversation. Every roadmap decision gets measured against that standard: whether it makes the advisor more capable, or quietly substitutes for one.
We’d rather say that here than three calls in. Most agencies talking to us this year aren’t unhappy with their platform; they’ve simply been handed a question it wasn’t built to answer.
ICHRA has crossed into your book, and your stack can’t do it without a second system.
It usually isn’t a strategy shift, just two clients who asked. But now you’re quoting in one tool and administering in another, reconciling the two by hand and explaining to a client why a qualifying life event lives somewhere different from their renewal. That’s the conversation we’re usually walking into, and the rest of this page is about whether we’re the right answer to it.
The ICHRA employee and the group employee are the same person in the same system, which means one eligibility rule set, one qualifying-event process, one place to look. Two platforms means two of each, reconciled by whoever has the spreadsheet open.
Failed carrier files, rate mismatches, and stalled elections show up in your queue with enough context to act on. The alternative, where the client finds it first, is a common way for a platform to damage an agency relationship.
Every election has a visible route and a full history, so “what happened to this enrollment” is something you look up in about ten seconds. An advisor who has to open a ticket to answer a client question is, for the length of that ticket, not the one advising.
Most platforms describe themselves as broker-friendly. The useful test is whether the business model would survive being otherwise. These three follow from how ours is put together.
We do benefits administration, and only that. There’s no bundled HR suite, onboarding module, or payroll add-on in here for your client to adopt casually and then call you about. We keep the scope narrow on purpose.
We hold no broker license, so we cannot write a group even if we wanted to. We don’t staff enrollment support inside your client’s business under our name, and we don’t earn commission on their coverage. We’re paid for software and nothing else. That is built into how the company is structured, which is why it isn’t something we could quietly revise later. When an employer reaches us directly, our first move is to ask who their broker is and bring them in. Only when there isn’t one do we involve a partner agency. Groups arrive through an agency and stay attached to one, because neither the role hierarchy nor the revenue model contains another route.
No house plan, no resale margin, and no carrier that quietly works better than the rest because of an arrangement you can’t see. You know the client and the market; the platform’s job is to execute the placement you chose and administer it correctly.
None of this was adopted for the ICHRA market. It’s what a decade of building benefits administration for agencies produces when there was never a direct-to-employer business waiting in the wings.
Operational differences, visible at renewal, at open enrollment closeout, and in the next client qualifying life event.
A prospect asks about ICHRA and you already know the answer. A client wants a hybrid strategy across employee classes, or a carrier gets added mid-year, and it’s all still the same system.
Renewals roll cleanly, and open enrollment closeout handles non-responders as a reviewable step instead of a silent batch. Carrier file failures surface before the client notices them, and qualifying life events process without a twenty-email thread.
Plan setup, benefit classes, contribution rules, eligibility, carrier connections — all yours to do, on your timeline, without filing a ticket or borrowing someone else’s implementation calendar.
Feature lists rarely decide these evaluations. Whatever the last platform did badly does, and after a decade of these conversations the questions arrive in roughly the same order.
You get that list before you sign. Sixty-nine group carriers are connected today, nine of them through integrations we built and maintain ourselves, and every carrier without an electronic path is labeled as manual up front — so nothing quietly turns back into a fax at renewal. See the full carrier list.
Plan designs, rates, and contribution rules carry forward into the next plan year, so a renewal means reviewing what changed, not rebuilding from scratch. And because the ACA rate math runs on our own infrastructure, the rate you quoted is the rate that renews.
Benefit classes are a property of eligibility and contribution, so one plan serves all of them. It’s the same mechanism whether you’re splitting hourly from salaried on a group plan or setting ICHRA contributions by class.
A short guided election with their real options, their real cost per pay period, and plain-language explanation of what each choice means — not a benefits education portal they have to be nagged into using. Where the decision gets hard, the path leads to you.
Weeks, not a quarter, and the work is mostly yours to sequence: agency setup, then a group at a time as their plan years allow. We don’t hold your calendar hostage to an implementation queue, and nobody has to migrate a live group mid-plan-year to get started.
Participation and waiver patterns, contribution and premium totals by class, election activity across the enrollment window, and the current state of every carrier submission. The reporting exists so you can walk into a renewal already knowing what the client is about to ask.
Either, for now. CMS and the SBA announced CHOICE — Custom Health Option and Individual Care Expense — as the new name in September, but the codification bills that would put it in statute haven’t passed, and the mechanics of the arrangement did not change. Your clients will start hearing CHOICE in the press while every carrier form still says ICHRA. We use both names in the platform for that reason, and we’ll follow whatever the market settles on.
No. We’re paid for software, never out of your compensation — no share of commission, no mapping fees on ancillary lines, no per-carrier surcharges. What you earn on vision and life stays what you earn on vision and life. What the software costs depends on the shape of your book, so we work that out against your own numbers on the first call. Any figure we posted here would be wrong for most agencies reading it. Where a distribution partner sets the price to the end client, they set it themselves, and our agreements require that it be disclosed to the employer in plain terms. We don’t dictate the number and we don’t allow it to be buried.
Platforms that resell a single submission network inherit that network’s gaps, which is how “connected” turns into a fax at renewal. Where a carrier sits outside our submission partners, we build and maintain the integration ourselves — and two of those nine can’t be reached through third-party networks at all.
See the full carrier listThere’s no self-serve signup for agencies, and that’s a choice about what we expect from the partners who distribute Abra.
Administering a group well takes staffing and follow-through. The conversation is where we establish that a partner can carry it — the licensing, the staffing at renewal, the willingness to answer an employee in January. Agencies and general agencies both come in this way.
We’ll work with whoever advises you today. If you don’t have a broker, or yours would rather not take this on, we’ll connect you with a partner agency we trust. Either way you end up with an advisor, because that is how the platform is meant to be used.