Lender Price, a provider of mortgage product and pricing technology, announced the next evolution of POD, its AI Pricing Optimization Dashboard: a human-supervised AI agent capability built to help deliver a new standard of pricing accuracy across the Lender Price Pricing Engine.

The pricing engine is the system lenders rely on to turn investor rate sheets into the rates and adjustments they ultimately quote to borrowers, which is why the accuracy of each update carries direct operational weight. POD AI Agents are designed to learn and execute the repeatable steps Lender Price pricing teams perform whenever investors publish changes to rates, loan-level price adjustments (LLPAs), product eligibility, programs, and pricing specials. LLPAs, the risk-based price adjustments investors apply to individual loans based on factors such as credit score and loan-to-value, directly shape the price on each loan quote. Working within defined guardrails and approval workflows, the agents can interpret incoming change information, prepare and validate system updates, identify exceptions, and route changes requiring judgment to Lender Price pricing professionals.

"In mortgage pricing, accuracy is not a feature — it is the foundation," said Dawar Alimi, CEO and Co-Founder of Lender Price. "POD is not about replacing the people who understand mortgage products, capital markets, and our clients' businesses. It is about giving them AI agents that can take on the painstaking, repeatable work so our experts can focus on controls, exceptions, and the decisions that deserve human expertise. The outcome we are pursuing is unparalleled pricing accuracy."

Built in Phases, Grounded in Pricing Operations

The POD initiative has progressed through multiple phases. Lender Price first applied AI to strengthen accuracy checks, helping teams surface potential pricing discrepancies sooner. It then expanded AI support for implementation work, organizing lender data, identifying gaps, and accelerating the onboarding of products, programs, eligibility rules, and investor-specific pricing.

The latest phase operationalizes that accumulated learning through AI agents trained on the step-by-step workflows used by Lender Price's own pricing team. Rather than allowing unchecked automation, POD applies controls that can include source data validation, rule-based guardrails, exception thresholds, change logs, testing, and human review before production release. The goal is to make updates faster and more consistent while maintaining clear accountability.

Expected Operational Impact

Lender Price expects POD to materially reduce opportunities for manual entry errors and improve the consistency of routine investor updates. Initial operating targets include:

• Up to a 90% reduction in manual touchpoints for eligible, repeatable rate sheet and pricing update workflows.
• Up to 75% faster preparation and validation of routine LLPA, pricing special, and program updates, with exceptions routed to an expert.
• At least 99.9% change traceability coverage for POD-processed updates, supported by source linkage, validation records, approvals, and audit logs.
• A meaningful reduction in preventable configuration errors — including accidental keystrokes, transposed values, or missed eligibility conditions — through structured extraction, validation rules, and controlled release workflows.

These figures are forward-looking operating targets, not historical customer performance results. Lender Price stated it will validate and publish measured outcomes as POD deployments mature. Those published results will give lenders a concrete benchmark for judging whether the targets hold in production.

Accuracy at Scale with Experts in Control

POD's design recognizes that mortgage pricing does not operate in a vacuum. Investor guidance can vary in format and complexity, and real-world updates often require contextual interpretation. The AI agents are intended to handle the high-volume, repeatable work, while escalation paths preserve human oversight for ambiguous guidance, unusual product structures, and changes that merit additional scrutiny.

The result is a pricing operation able to respond more quickly to investor changes without sacrificing the controls lenders expect. By reducing repetitive manual work, POD enables Lender Price's pricing experts to devote more time to quality assurance, edge cases, strategy, and client-specific needs.

"The future of POD is a more accurate, more resilient pricing operation," Alimi added. "Our people essential. AI agents make their expertise more powerful by helping ensure that every update — from an LLPA change to a pricing special or eligibility rule — is handled with discipline, visibility, and speed."

POD is part of Lender Price's broader AI strategy, which applies AI where it can improve implementation efficiency, strengthen pricing quality controls, and help lenders move with greater confidence in a fast-changing market. The human-supervised structure of POD also reflects a wider pattern in financial technology, where AI agents are increasingly deployed under human review rather than full autonomy in regulated areas of financial services.