← All products
LMLoss mitigation

LM Platform

AI-native loss mitigation for mortgage servicers.

LM Platform lets mortgage servicers manage the entire lifecycle of every workout under its investor’s rules — without the manual handoffs, spreadsheets or vendor development cycles legacy default-servicing systems require. Built on the Pulsar core, it runs cases through configurable, agency-specific workflows instead of one-size-fits-all queues.

Talk to us See capabilities
LM Platform work dashboard
19+FHA, FNMA & FHLMC programs prebuilt
8staged workflow from intake to closing
9modification step types in Mod Calc

Overview

At a glance

01

Configurable workout workflow engine

Multi-stage cases — staged, SLA-tracked and team-owned — published and versioned without custom code.

02

Agency-specific modification programs

19+ prebuilt FHA, FNMA and FHLMC workflows encoding real eligibility and stage logic.

03

AI case-work agents

A stage-level copilot and borrower-message assist, with underwriting, QC and document-extraction assistants on the same infrastructure.

04

Continuity-first auto-assignment

Cases stay with the same team member across the stages their team owns.

05

Mod Calc

A bundled, no-code modification waterfall calculator that returns a full audit trace, not a single number.

Capabilities

What makes it different

Seven capabilities that compound — six inside the case-management core, plus Mod Calc’s no-code calculation engine bundled alongside it.

01

Configurable Workout Workflow Engine

Every workout runs through a staged workflow — intake, document collection, underwriting, second-look review, investor approval, trial payment, final calculation and closing — each stage with its own SLA timer, team ownership and document checklist. Workflows are configured and published like software, so a new program or an SLA change is a configuration edit, not a vendor development ticket.

Why it matters
  • Each stage’s document tracking is auditable in place — recording, counter-signature and delivery are timestamped
  • SLA timers and team ownership are configurable per stage
  • The same engine every Pulsar-based product runs on, not a bespoke case-management build
Configurable Workout Workflow Engine
02

Agency-Specific Modification Programs

19+ prebuilt program configurations span FHA, FNMA and FHLMC modification, forbearance, repayment, short sale and deed-in-lieu workflows — each with its own stage sequence, SLA hours, eligibility rules and linked calculator. Activating a new program or retiring an old one is an admin action, not a code release.

Why it matters
  • Real agency program logic ships out of the box
  • Eligibility is evaluated per program, so ineligible paths are caught before they waste underwriting time
  • New or amended agency guidance becomes a configuration, not a multi-week project
Agency-Specific Modification Programs
03

AI Case-Work Agents

A stage-level AI copilot and a borrower-message assistant are live in the case workbench, grounded in a knowledge base of program guidelines. An underwriting assistant, an automated QC engine and a document-extraction agent are built on the same model-agnostic infrastructure — routed to a local model first, with automatic cloud fallback — ready to plug into any stage.

Why it matters
  • Model-agnostic routing means no capability depends on one AI vendor’s uptime
  • Every suggestion carries a confidence flag, so a processor sees what the AI is unsure about
  • Additional agents are ready for clients who need more automation and compliance coverage
AI Case-Work Agents
04

Continuity-First Auto-Assignment

Cases auto-assign to the least-loaded eligible member of the owning team on every stage transition, keeping the same processor across the stages their team owns. Managers see live active-case counts by team, pause a member’s eligibility for leave in one click, and bulk-reassign cases the moment someone goes out of office.

Why it matters
  • Work is never stuck on someone who is out
  • Context stays with the same processor across a case
  • Every pause, resume and bulk reassignment is audit-logged
Continuity-First Auto-Assignment
05

Analytics & Operational Dashboards

Live, no-code-configurable dashboards show total and active case counts, SLA compliance by cohort, pipeline volume by month and per-team capacity — all queried directly from production case data, not a stale export. A Dashboard Designer lets an admin scope a new dashboard to a program type or queue without a development request.

Why it matters
  • SLA compliance and breaches are visible in real time, by program and by team
  • Team Capacity shows which queue is overloaded before it becomes a missed deadline
  • New dashboards are a configuration exercise, not an engineering backlog item
Analytics & Operational Dashboards
06

Borrower Self-Service Portal

A mobile-first borrower experience tracks a case through six clear steps — hardship survey, financial info, review & sign, program match, accept plan, complete — with a dedicated point of contact, required-document status and e-signature built into the flow. Borrowers see exactly where their application stands without a phone call.

Why it matters
  • Fewer inbound status calls — borrowers self-serve against a visible step tracker
  • Document status (approved, pending, required) is visible in real time
  • E-signature and upload are native to the portal, not a separate emailed PDF
Borrower Self-Service Portal
07

Mod Calc — No-Code Modification Waterfall Engine

Bundled alongside LM Platform, Mod Calc lets a compliance analyst assemble a modification waterfall — capitalize arrears, rate reduction, term extension, forbearance, deferral, partial claim — by dragging preset steps into order and setting an affordability target. Every run produces a before/after summary and a step-by-step audit trace showing which steps applied, which were skipped, and why.

Why it matters
  • A new or changed agency program is configuration, not a development ticket
  • Every calculation is fully auditable — the trace shows how the final terms were reached
  • Handles ARM and simple-interest loans correctly by using the actual entered payment
Mod Calc — No-Code Modification Waterfall Engine

Built in

Also built into the platform

LM Platform ships the operational infrastructure a servicer needs on day one.

Batch Jobs

Scheduled and on-demand background processing.

Single Sign-On

Real OIDC integration with auto-provisioning and attribute mapping.

Two-Factor Authentication

TOTP-based 2FA with backup codes on every account.

Team & User Management

Role-based access control and organization hierarchy.

Security Events & Dashboard

Full audit trail, queryable by event type, user or email.

E-Signature Integration

Self-hosted e-signature with primary and co-borrower signers.

Rules Engine

Visual, no-code rule sets evaluated against live case data.

Dashboard Designer

No-code dashboard configuration scoped to any program or queue.

Who it’s for

Who it’s for

Servicers absorbing rising delinquency and modification volume on the same legacy systems and headcount they had a decade ago.

Mid-market bank servicersRegional and community bank servicing arms managing 250K–1M loans, balancing compliance risk against thin loss-mitigation staff.
Non-bank & independent servicersStandalone servicing shops scaling loan volume faster than their loss-mitigation headcount.
Specialty servicers & subservicersFirms servicing distressed and specialty portfolios on investors’ behalf, where modification volume is the core business.
Credit unionsMember-owned servicers under the same CFPB loss-mitigation obligations, with the smallest teams to absorb them.

Architecture

Under the hood

How LM Platform is built

Configuration-first and agency-aware — every program, AI agent and calculation runs through the same auditable case record.

01PresentationConsole, workbench, borrower self-service
Admin ConsoleRole-gated admin console
Case WorkbenchFull case record, stage workflow, AI copilot
Borrower PortalMobile-first, step-tracked application experience
02Case & WorkflowAgency-specific workout workflows
Workflow EngineStage state machine and agency program rules
Agency Programs19+ FHA, FNMA, FHLMC workflows stored as versioned config
Continuity-First AssignmentRe-evaluated on every stage transition
03AI Case-WorkModel-agnostic agents
Stage Copilot & Text AssistLive today — local-model first, cloud fallback, RAG-grounded
Underwriting AssistantBuilt, ready to route in
QC & Doc ExtractionBuilt, ready to route in
Knowledge Base (RAG)Hybrid search grounded into every response
04Data & AuditDurable, audit-logged
PostgreSQLPrimary case and workflow data
Job QueueBackground batch processing and scheduling
Audit & Security EventsFull audit trail and security-event logging
05IntegrationConnected to your stack
Mod CalcCertified modification waterfall engine over REST
SSO / Identity ProvidersOIDC per organization
Investor RoutingFannie Mae SMDU and Freddie Mac Resolve, integration-ready

How Mod Calc is built

No-code and audit-first — every waterfall step is configuration, and every run produces a complete before/after trace.

01ConfigurationZero dev cycles
BuilderDrag-and-drop, no code, no deploy cycle
Agency PresetsFNMA Flex, FHLMC Flex, FHA, VA, USDA — one click to pre-fill
Calculator LibraryNamed, saved calculators recalled by any program
02ExecutionOne engine, identical logic everywhere
Waterfall Engine9 step types run against an affordability target
Target EvaluatorPayment reduction %, DTI, PITI, affordable payment
Step-by-Step Audit TraceEvery step logged before/after, applied or skipped
03AccessTwo ways in
Interactive TesterRun any saved calculator against real loan inputs
REST APIAPI-key gated; LM Platform calls it in real time

Keep exploring

Other products on Pulsar

P PulsarPlatform PSP Pulsar Servicing PlatformPSP AI Ai.DLCSDLC governance M MiraWealth & lead-gen CS Pulsar CSComing soon

Talk to us about LM Platform.

Get in touch