Laudio

What the platform sells first is integration and workflow; the models sit on top. Laudio pulls a frontline leader's data from human resources, time and attendance, scheduling, email, the learning management system and real-time EHR patient feeds into one hub for recognition, check-ins, rounding, attendance, performance and scheduling work, and that consolidation would carry value with no model in it. The FAQ says the company has used AI and the latest large language models since its founding: AI-driven recommendations surface which leader action to take with which team member, a Turnover Risk Score uses machine learning and predictive analytics to flag team members at higher risk of leaving, and Performance Insights (October 2025) generates AI summaries of performance data for reviews. A staffing module uses an optimization engine to draft schedules.

The learned components are found on the company's pages and matter to the product, but they are features of a leader workflow platform rather than the thing sold.

The system recommends and flags, and a leader acts; the basis of the flags is the missing part. The company's pages describe AI-driven recommendations for leader actions, system-generated flags for burnout and turnover risk so leaders can intervene early, AI summaries that support performance conversations, and a scheduling optimizer whose drafts are pushed back to UKG or Kronos only once a manager approves them. Nothing published suggests the system contacts, disciplines or evaluates an employee on its own.

What is not published is what raises a turnover or burnout flag, whether the leader sees the contributing factors, and whether an employee is ever told a flag exists. A nudge system still decides who gets a manager's attention and who does not.

The approach is described in general terms and nothing is identified. The FAQ says the platform uses AI and the latest large language models, that the Turnover Risk Score uses machine learning and predictive analytics, and that Performance Insights generates AI summaries around pillars tailored to each organization; the staffing page describes an optimization engine. No model, model family or provider is named, the inputs to the risk score and the weighting behind performance summaries are not described, and no version or update practice is published.

Performance Insights applies a summary to a named employee that enters a review conversation, so its inputs matter.

The infrastructure is named and the model layer is not. The public security page names Amazon Web Services as the host, AWS Key Management Service for encryption keys and Vanta for compliance monitoring, and the FAQ says the platform uses the latest large language models. The provider of those models is not named, and the company publishes no subprocessor list, though the trust portal may hold one. The Ascend Learning privacy policy adds Ascend's own brands as recipients, with information linked across them.

Two datasets travel this chain: real-time patient data from EHR feeds and a behavioral record on named employees.

Named health systems report dated outcomes on one consistent measure, all through the company. The FAQ and client pages report that UNC Health reduced RN turnover by 20 percent, saved $5.4 million a year and reported more capacity for 94 percent of managers; that Novant Health cut nurse turnover by 15 points, saved $90 million in agency costs and avoided $4.8 million in replacement costs; that El Camino Health reduced RN turnover by 7 points and avoided $841,000 in replacement costs; and that Nebraska Medicine cut first-year RN turnover by almost half. The press page links independent coverage, including Healthcare IT News on UNC Health (August 2025) and North Mississippi Health Services saving $2.7 million (March 2026), and the company publishes research with AONL on nurse burnout.

None of the results has a control group or adjusts for other retention programs running at the same time, and none is independently evaluated, so the platform's contribution is not isolated.

One categorical commitment is published beside grants that cut against it. The Ascend Learning global privacy policy (effective September 2026), which Laudio's site links as its privacy policy, says personal data is not used to train large language models and is not sold for that purpose. The same policy says information may be disclosed to generate reports and analytics for researchers and third parties in the fields Ascend supports, may be shared among Ascend's brands and linked across them to recognize the same individual, and is kept as long as necessary for its purposes, with no period stated. Laudio publishes research reports built from its customers' workforce data, such as its nurse burnout analysis with AONL.

The platform holds two sensitive datasets: real-time patient movement data from EHR feeds for rounding, and attendance, overtime, recognition, peer input and performance records on named employees, from which it derives turnover risk. Nothing says how long derived risk scores and AI summaries are kept or whether they train the risk model. Safety engineering, meaning how a wrong or harmful AI summary is caught and handled, is not published.

Business associate terms are stated conditionally and HIPAA compliance is backed by an outside audit, with the agreement unpublished. The platform takes real-time patient data from EHRs for rounding, so HIPAA applies. The public security page says Laudio has cleared an independent HIPAA audit and offers the documentation on request through its trust portal, and the FAQ says it complies fully with HIPAA. The Ascend Learning privacy policy that governs the site says that where a customer's contract requires it, Ascend signs a business associate agreement and adheres to its obligations, and that for end user data Ascend acts as the customer's processor.

No agreement template is published, and nothing says which data classes the agreement covers or whether the employee records, which are not protected health information, sit inside the same terms.

Three attestations are named, and the documents are available on request rather than open. The public security page says Laudio holds a SOC 2 Type II report and a TX-RAMP Level 2 certification, a Texas state program with an assessment process, and has cleared an independent HIPAA audit; it describes third-party penetration tests that evaluate source code and the deployed environment, static analysis, and continuous monitoring through Vanta.

It asks buyers to request the documentation through trust.laudio.com, a Vanta portal, and nothing public shows whether the reports are self serve or released on approval; the site gives no report date, period or scope.

No device claim is made and none applies: the platform supports managers' work with their teams, and its patient data feed drives rounding workflows, not clinical decisions. What governs it is employment law and, for patient data, HIPAA, which the company addresses on its security page. Performance Insights produces AI summaries used in performance conversations, which brings it within emerging state rules on automated decisions about employees and, at unionized systems, within bargaining over how algorithms enter performance management; nothing published addresses either.

Employment law is named here as what governs the product; it is not a device pathway. Buyers with represented staff should raise the performance summaries with labor relations before deployment.

Policy language is published and no evaluation is. The FAQ says the company uses AI to strengthen, not replace, the human connections of leadership, and the Ascend Learning privacy policy that governs the site says personal data is not used to train large language models. Nothing describes how the Turnover Risk Score or Performance Insights is tested, whether flags or summaries differ across groups of employees, or any bias review.

The risk is specific. The turnover and burnout signals draw on attendance, overtime and work patterns, which vary with caregiving responsibilities, disability, seniority and shift type, and performance summaries draw on recognition and peer input, which are not distributed neutrally across race and gender. A flag that marks an employee as a turnover risk can change how a manager treats her.

The outputs reach a leader who decides, and nothing stands behind them. Recommendations, turnover risk flags and Performance Insights summaries are shown to frontline leaders, who choose what to do; Performance Insights organizes the underlying performance data alongside the summaries, so a leader can check a summary against its sources, and the staffing module pushes schedules back to UKG or Kronos only after a manager approves them. That review by the leader is the only mechanism published.

The employee the flag or summary is about has no stated route to see it, understand its inputs or contest it; the Ascend Learning privacy policy treats the employer as controller of end user data, so access and correction run through the employer, and they reach personal data rather than the model's output. No accuracy figure for the Turnover Risk Score and no remediation commitment are published. A turnover flag can also cause what it predicts, if a manager invests less in someone marked as likely to leave.

Integration is the product, and one integration is named with its direction. The Staffing and Scheduling page describes two-way integration with UKG and Kronos: the platform pulls scheduling and self-scheduling data, drafts a schedule with an optimization engine and pushes the finalized schedule back once a manager approves it. The FAQ lists integrations with EHRs (real-time patient data for rounding), time and attendance, HRIS and HCM systems, email and learning management systems, and a UNC Health nurse manager describes the platform as integrated with the system's timekeeping, scheduling portal and HR system.

No EHR is named, so the clinical feed cannot be matched to a buyer's record system, and apart from UKG and Kronos no HR, timekeeping or learning system is named.

A single hosted service on a named cloud, with location implied rather than committed. The public security page says the application is hosted on Amazon Web Services, with data encrypted at rest and in transit and keys managed in AWS Key Management Service; no region, residency commitment or tenancy model is stated. The company also holds a TX-RAMP Level 2 certification, which requires a documented system boundary; that document is not published.

Laudio was acquired by Ascend Learning in 2025, and the Ascend privacy policy that now governs the site allows information to be shared and linked across Ascend's education, assessment and workforce brands. A nurse may appear in Ascend's training businesses and in her employer's Laudio data.

No price, unit or charging basis is published; every path leads to a demo request. The FAQ says implementation varies with an organization's size and the features deployed, usually from a few weeks to a couple of months, and client results are framed in avoided turnover and agency cost (UNC Health $5.4 million a year, Novant Health $90 million in agency costs), which sets the purchase against savings without stating the cost. What the site does make clear is that the platform is sold to health systems, is modular (recognition, rounding, performance, staffing) and scales with the organization.

Coverage is named and supported for acute care hospital systems. The client pages name UNC Health, Novant Health, El Camino Health, Nebraska Medicine, Boston Medical Center, North Mississippi Health Services, Memorial Hermann and Northwell Health, with results for several, and the product is aimed at nursing executives, HR executives and frontline leaders across departments, not nursing alone.

Nothing is published for ambulatory, post-acute, senior living, home care or behavioral health settings, and the results reported are for nurse turnover rather than other roles.