Questionnaire Maintenance
Administrators write the wizard's questions — one sentence and up to four answers each — per Access Control in the configuration console, and change them without a developer.
Smarter Benefits Decisions — Right When Employees Need Them
The personal benefits advisor your workforce deserves.
IntelliSuggest is a short wizard inside enrollment. The employee says who the coverage is for and answers a few questions the benefits team wrote; the model reads their profile, dependents, medications and doctors, asks what else matters, and ranks the plans already on their offer — best plan first, with the reasons. Adding a result uses the same cart as every other election.
Only plans on the employee's offerReasons on every rankInside enrollmentRecommends · never elects
Understanding Each Employee — Without Guesswork
"Most employees re-elect last year's plan because the alternative is a spreadsheet and a guess. Decision tools bolted onto enrollment ask them to retype what the employer already knows, then recommend a plan they may not be offered. IntelliSuggest starts from the record and the offer: a few questions the benefits team wrote, the profile the platform already holds, and a ranking limited to the plans this person can actually put in the cart."
Intro & CoverageWho is the coverage for?Stored QuestionsWritten by your benefits teamInsights & Follow-upsWhat else matters to you?Ranked OffersBest plan, alternatives, reasonsCart, Compare, AskThe same cart as every election
1offer list the model may choose from — the plans already calculated for this employee and this coverage, nothing outside it
4answers at most per stored question, single choice, one question at a time — written and maintained by administrators
~8follow-up questions proposed from the employee's context, plus a free sentence of their own
0elections made by the wizard — Add to cart is a draft on the normal enrollment cart; the employee signs on Review & Sign
Working flow
Who Clicks What, in Order
The wizard as it runs today — who does what, in order, and the rule the platform enforces at each step.
| # | Who | What they do | What the platform enforces |
|---|---|---|---|
| 1 | Administrator | Writes the questions the wizard will ask — one sentence and up to four answers each — in the questionnaire screen of the configuration console. | Questions are read by the wizard; the employee's answers are never stored — they live only for the session. |
| 2 | Employee | Opens enrollment; the wizard opens when plan rows exist (automatically once, or from the IntelliSuggest button). | No offer rows, no wizard — a warning explains that plans are not yet available. |
| 3 | Employee | Reads the how-it-works intro and picks who the coverage is for — employee only, employee + spouse, family and so on. | The coverage choice decides which dependents the model may consider and which offers are eligible. |
| 4 | Employee | Answers the stored questions one at a time, single choice. | Answer texts marked non-mandatory are hidden; progress moves through the middle of the bar. |
| 5 | Platform · model | After the last answer, two calls run together: one "My…" insight sentence per answer, and about eight follow-up questions prioritised from the employee's profile, dependents, medications and physicians. | Priorities always start with cost, out-of-pocket maximum and preventive care; pharmacy is added for a heavier medication list; mental health and telehealth are always offered; the list is cut at eight. |
| 6 | Employee | Picks the follow-ups that matter and, optionally, types one sentence of their own. | The selections and the sentence become the employee's query for the ranking; three rotating status lines show while plans load. |
| 7 | Platform · model | Loads the latest offer for this person and coverage, collapses it to one row per plan and option, and asks the model for a best plan and alternatives — with a score, why it matches, reasoning, a recommendation summary and an analysis block. | A code not on the offer is rejected and the call retries; the screen keeps a plan only if the same plan, option and coverage is already on the enrollment rows it shows. |
| 8 | Employee | Reads the results — best plan first — zooms, compares, or asks Athena from the wizard header. | Cost figures on the card are copied from the offer row, never from the model's arithmetic. |
| 9 | Employee | Adds a plan to the cart (with the dependent selection the cart already uses) and continues enrollment. | The cart is a draft; the election is made on Review & Sign by the enrollment engine with its eligibility rules and audit trail. |
One Employee, Five Screens to a Confident Choice
Follow Jane Doe — spouse, one child, two prescriptions, three doctors she wants to keep — from the intro screen to a plan in the cart.
Who is the coverage for? That decides everything after it
Jane's enrollment has offer rows, so the wizard opens. The intro explains the steps in plain words — the benefits team can add its own intro copy — and Accept moves on. The coverage screen lists the options her plan offers: employee only, employee + spouse, employee + children, family. She picks family. From this point the platform knows which dependents count (those whose relationship the coverage allows) and which offers are eligible (those calculated for that coverage).
What the screen does
- Coverage options come from the enrollment's own dependent-coverage list — the same codes the cart uses
- Employee-only is recognised whether the code or the label says so
- The choice is held in the wizard only — nothing is written to the record
- Back returns to the intro; progress shows 15%
Why it matters
A recommendation for the wrong household is a wrong recommendation.
Family selects the spouse and child on file; employee-only leaves them out of every later call.
A few questions — written by your benefits team, not by a vendor
The wizard shows one stored question at a time with up to four answers, single choice: how often the family sees a doctor, whether a procedure is planned, how they feel about a higher deductible in exchange for a lower premium, whether keeping current doctors matters. Administrators write and maintain these in the configuration console, per Access Control. Jane's answers are held in the wizard for the session and sent with the two calls that follow — they are never stored as a profile.
What the screen does
- One question per screen; Back returns to the previous question or to coverage
- Only answer texts that exist and are not marked non-mandatory are shown
- Questions are loaded when enrollment opens, so the wizard never waits on them
- Administrators add, edit or delete a question without a developer
Why it matters
The questions reflect your plans and your workforce, in your words.
A question about a procedure this year changes the ranking more than any demographic field.
"My…" insights, then the questions that matter to you
When Jane answers the last question, two things happen at once. The model turns each answer into one short insight sentence that starts with "My…" — "My family uses regular prescriptions and sees a specialist now and then" — so she can see what the wizard understood. And it proposes about eight follow-up questions built from her context: cost, out-of-pocket maximum and preventive care always come first; pharmacy is added because two medications are on file; network and family because she wants to keep three doctors and covers a child; mental health and telehealth are always offered. She ticks the ones that matter and can type a sentence of her own.
What the screen does
- One insight per answer, at most fifteen words, in plain language — no plan acronyms unless the employee used them
- Follow-ups are chips the employee can multi-select; a custom sentence is welcome
- Priorities come from the record — medications, physicians, dependents, profile — not from a quiz
- Either call failing keeps the employee on the questions with a clear message; nothing half-finished is shown
Why it matters
The employee corrects the wizard before it ranks, not after.
Selected follow-ups and the sentence become the query the ranking is asked to honour.
Tailored, transparent plan guidance — from the offer, never outside it
The platform loads the latest offer calculated for Jane, keeps the plans for family coverage, collapses them to one row per plan and option, and sends the model the offer's own figures — employee cost, deductible, out-of-pocket maximum, coinsurance — together with her profile, dependents, medications, physicians, answers, insights and query. The model returns a best plan and alternatives, each with a score, why it matches and its reasoning; the best plan also carries a recommendation summary, an analysis across cost, network, coverage and priorities, an overall rating and a confidence. Every cost on the card is copied from the offer row.
What the screen does
- Best plan first, then alternatives, each with the model's reasons in plain sentences
- A plan code that is not on the offer is rejected and the request is retried — up to three times
- The screen keeps a result only when the same plan, option and coverage is on the rows already shown in enrollment
- Zoom widens the dialog; "No plans matched your criteria" is shown honestly when nothing qualifies
Why it matters
Employees follow advice they can argue with — and can actually elect.
A stand-alone tool can recommend a plan you are not offered; this one cannot.
Add to cart, compare, or ask Athena — the wizard never elects
Jane taps Add to cart on HDHP Plus. The wizard hands the plan to the same health-plan cart every other election uses — with the dependent selection dialog and the "already in cart" check that cart already has. The cart is a draft; the election is made on Review & Sign by the enrollment engine with its eligibility rules and audit trail. Compare opens the enrollment's own comparison. "Ask Athena" in the wizard header opens the employee chatbot for anything the ranking raised — what a deductible means for her family, whether a birth reopens her choices — answered from her records and the plan documents.
What the screen does
- Add to cart, compare and dependent selection are the enrollment cart's dialogs, not a second cart
- Closing the wizard clears its answers, insights and results; nothing lingers on the record
- "Ask Athena" opens the same chatbot the employee uses everywhere in the portal
- The plan tile's own help button asks one field question with the numbers reloaded from the offer
Why it matters
One cart, one signature, one audit trail — the recommendation never becomes a shortcut around them.
Athena explains; the wizard ranks; the enrollment engine elects.
Ground rules
Rules of the Wizard
What the employee can rely on and what an auditor can check — from the wizard as it runs, not from a brochure.
| Rule | How it behaves | Why it matters |
|---|---|---|
| Only the offer | The model chooses from the plans already calculated for this employee and coverage; a code it invents is rejected and the request retried; the screen keeps a plan only if it is on the rows shown. | No employee is ever recommended a plan they cannot elect. |
| Numbers from the offer row | Employee cost, deductible, out-of-pocket maximum and coinsurance on every card are copied from the offer, not produced by the model. | The figure the employee reads is the figure the cart will charge. |
| Never elects | Add to cart is the enrollment cart; the election is made on Review & Sign by the enrollment engine. | One cart, one signature, one audit trail. |
| Coverage first | The coverage the employee picks filters both the dependents the model may consider and the offers it may rank. | A family recommendation is built for the family on file. |
| Context from the record | Profile, employment, dependents, medications and physicians are loaded on the server; the employee is not asked to retype them. | Fewer questions, better answers, no second profile to keep in sync. |
| Questions are yours | Stored questions and answers are maintained by administrators per Access Control; the wizard only reads them. | The questions fit your plans and your people. |
| Answers are not kept | The employee's answers, insights, follow-ups and results live in the wizard for the session and are cleared when it closes. | No preference profile accumulates behind the employee's back. |
| Reasons on every rank | Best plan and each alternative carry a score, why it matches and reasoning; the best plan adds a summary, an analysis block and a confidence. | An employee can disagree with a reason and change an answer. |
| Athena is one tap away | The wizard header opens the employee chatbot; the plan tile's help button asks a field question with numbers reloaded from the offer. | Questions the ranking raises are answered from the same records, with a source. |
Core Capabilities That Define BR IntelliSuggest
Six pieces — two for administrators, four for employees — that together make a recommendation the employee can trust and elect.
Plan Feature Library
Administrators can extract the key features of a plan from its brochure or summary and store them for the plan category — a maintained reference for plan comparisons and employee questions.
Personalized Recommendations
Best plan and alternatives chosen from the employee's own offer, using their profile, dependents, medications, physicians, answers and follow-ups — with a score and the reasons for each.
Insight & Follow-up Questions
One plain "My…" sentence per answer so the employee sees what was understood, and about eight follow-up questions prioritised from their context, plus a sentence of their own.
Conversational Interface
"Ask Athena" on the wizard and the help button on every plan tile open the employee chatbot, where a deductible, a network or a life-event window is explained from the employee's own figures with a source.
Platform-Native Integration
Runs inside the enrollment the employee is already in — SAP SuccessFactors or any synced HRIS — and hands results to the enrollment cart, so the recommendation and the election share one screen and one audit trail.
Stand-Alone Decision Support vs. IntelliSuggest
Stand-alone decision tools are good at the conversation. They are far from the record, far from the offer, and far from the cart.
| Moment | Stand-alone decision-support tool | BR IntelliSuggest |
|---|---|---|
| Getting started | A long quiz about family, income and health. | A few questions your team wrote; profile, dependents, medications and doctors come from the record. |
| Which plans it can recommend | Any plan in its catalog — sometimes one the employee is not offered. | Only the employee's own offer, for the coverage they chose; anything else is rejected. |
| The numbers | Estimates the tool computed. | The offer's figures — employee cost, deductible, out-of-pocket maximum, coinsurance — copied onto the card. |
| The recommendation | "We suggest Plan B" with a generic explanation. | A best plan and alternatives, each with a score, why it matches and reasoning; a summary, an analysis and a confidence on the best. |
| Electing | Copy the plan name into a different portal. | Add to cart — the same cart, dependents dialog and Review & Sign as every election. |
| Questions afterwards | A help article. | Ask Athena from the wizard, answered from the employee's records and plan documents with a source. |
| What it keeps about the employee | A preference profile on the vendor's cloud. | Nothing — answers and results are cleared when the wizard closes. |
Athena beside IntelliSuggest
The Ranking Answers "Which?" — Athena Answers "Why?"
A ranked list is the start of a conversation. "Ask Athena" in the wizard header opens the employee chatbot, where the questions a ranking raises — what the deductible really means, whether a doctor is in the network, what a birth would change — are answered from the employee's own records and the plan documents, with a source.
- "Why is the HMO third?" — the physician that is out of network, named from the record
- "What does out-of-pocket maximum mean for us?" — explained with the family's own figures, cited to the plan summary
- "Can I switch plans if the baby comes?" — the life-event window, and a Pending Action to come back to
- The "i" on a plan tile — one field question with the numbers reloaded from the offer, never trusted from the browser
- For administrators: Athena for Admins explains the questionnaire and enrollment screens from approved help text
Why is HMO Select ranked third? It's the cheapest.
Athena · from your recordsIts premium is the lowest, but one of the three physicians on your record — Dr. Lin — is not in the HMO Select network, and out-of-network care isn't covered on an HMO except in emergencies. You told the wizard keeping your doctors matters, so the ranking counted it. If you'd switch doctors, say so and re-run the wizard.
Show the network listRe-run IntelliSuggest
We might have a baby next year. Should I pick PPO now or wait?
Athena · from your documentsBoth work. A birth is a qualifying life event, so you'd have 31 days after the birth to move from HDHP Plus to PPO Standard mid-year. Picking PPO now costs about $75 more per paycheck if the baby doesn't arrive; waiting means the HDHP deductible applies to prenatal care before you switch. Want a Pending Action so you're reminded when the window opens?
Add reminderCompare the two plans
Ask about these plans…
Why Organizations Choose BR IntelliSuggest
Better Plan Migration
3x more employees migrate to cost-efficient plans like HDHPs — the live estimate — because the recommendation comes with reasons they can read, not a brochure.
Nothing Outside the Offer
Every plan shown is one the employee can elect for the coverage they chose — the ranking is guarded on the server and again on the screen.
Higher Employee Confidence
Clear explanations an employee can repeat to a spouse, insight sentences that show what was understood, and Athena for the questions in between.
Questions in Your Words
The benefits team writes and maintains the wizard's questions per Access Control — about your plans, your workforce, this year's concerns.
Inside the Enrollment You Run
Seamless SAP integration — and every other HRIS Benefits Reimagined syncs — with HIPAA-aligned handling; the ranking and the cart on one screen, in one tenant.
One Cart, One Signature
Add to cart is the enrollment cart; Review & Sign is the election. A recommendation never becomes a shortcut around eligibility rules or the audit trail.
Personalized Benefits Decisions. Enterprise Intelligence.
IntelliSuggest is a modal in the employee's enrollment and two screens in the administrator's configuration console. Employees meet it when their offer is ready — automatically the first time, then from the IntelliSuggest button — and again on every plan tile's help button. Administrators maintain the questionnaire per Access Control and, where they choose, extract and store the key features of each plan from its documents.
- Employee: Open Enrollment and life events — the wizard opens whenever offer rows exist; results go to the cart the employee is already using
- Employee: plan-tile field guidance — one question about cost, deductible or similar, answered with figures reloaded from the offer and the approved field help
- Employee: Athena — from the wizard header, the same chatbot as everywhere in the portal
- Admin: questionnaires — question, up to four answers, per Access Control; add, edit, delete
- Admin: plan feature library — extract key features from a plan document and save them for the category
Four Stories, End to End
Four moments traced through the flow above — including the two where the guard, not the model, has the last word.
The family that kept its doctors
Jane picks Family, answers five questions, and ticks "my three doctors in network". The model loads her physicians from the record, finds that one is out of the HMO's network, and ranks HDHP Plus first with that reason written out. She adds it to the cart; the election is signed on Review & Sign.
The employee with no offer yet
Marcus opens enrollment before his offer has been calculated. There are no plan rows, so instead of the wizard he sees a clear warning that plans are not yet available — the wizard never ranks against an empty or stale list.
The model's invented plan
On one run the model returns a plan code that is not on the offer. The server rejects it — "recommended plan not found" — and retries; the second answer is on the offer and passes. On screen the plan is kept only because the same plan, option and coverage is already among the enrollment rows shown.
The question the ranking raised
The results say PPO Standard is safer if a baby is likely. Jane taps Ask Athena, asks whether she can switch mid-year, and gets the life-event window from the plan documents with a Pending Action to come back to. Athena explained; the wizard ranked; the cart will elect.
See a Recommendation Built for One of Your Employees
Bring your plan lineup and rates, five questions you would ask your workforce, and three anonymized employee profiles. We'll load the questionnaire, walk the wizard as each employee, and show the ranking, the reasons and the guard that keeps it inside the offer.
Only the employee's offerReasons on every rankNever electsAnswers not storedHIPAA-aligned
BR IntelliSuggest FAQs
An AI-powered benefits recommendation tool that helps employees choose the right plans based on data, preferences, and eligibility. It is a short wizard inside the enrollment screen: the employee says who the coverage is for, answers a few questions the benefits team wrote, adds anything on their mind, and gets a best plan and alternatives chosen from the offers already calculated for them — each with the reasons.
Two things are combined. What the employee tells the wizard — coverage choice, answers, follow-up concerns — and what the platform already knows: profile and employment, dependents allowed under that coverage, medications and physicians on file. The ranking uses the offer's own cost figures (employee cost, deductible, out-of-pocket maximum, coinsurance), so the numbers shown are the plan's numbers, not the model's arithmetic.
No. Results can be added to the same enrollment cart every other plan uses; the election is made there, on the enrollment engine with its rules and audit trail. The wizard has no enroll action of its own.
No. It may only choose from the plans on that employee's current offer for the coverage they picked. A plan code the model invents is rejected on the server, and the screen keeps a recommendation only when the same plan, option and coverage is already on the enrollment rows in front of the employee.
Benefits administrators, in the questionnaire screen of the configuration console — one question at a time with up to four answers, per Access Control. The wizard reads them; it does not store the employee's answers, which live only for the length of the session.
After the last stored answer the model writes one short "My…" insight per answer and proposes about eight follow-up questions, prioritised from the employee's context: cost, out-of-pocket maximum and preventive care always; pharmacy when the medication list is heavy; network, family, specialist, emergency, maternity, payroll cost, mental health and telehealth as the profile suggests. The employee picks the ones that matter or types their own sentence.
It is fully SAP-native, syncing real-time data from payroll and HRIS inside SAP SuccessFactors — and the same wizard runs for Workday, UKG, ADP, Dayforce, Paycom or any HRIS Benefits Reimagined syncs, because it reads the platform's own employee record and offer.
"Ask Athena" in the wizard header opens the employee chatbot, where questions about a deductible, a network or a life-event window are answered from the employee's own records and the plan documents, with a source. Athena explains; the wizard ranks; the cart elects.
Explore Benefits Reimagined Today
A recommendation that starts from the record, stays inside the offer, explains itself, and ends one tap from the cart — decision support that belongs inside enrollment.