Optimize Every Consumer Interaction

Dynamic Scoring Models

Powered by your own de-identified data on consumer behavior, Attunely’s machine learning platform determines the most effective outreach tactics through every phase of the servicing, delinquency, and recovery life cycle. Our models are customized to deliver insights that maximize your outreach strategies.

Propensity-to-Pay Model

Predict whether payment is likely from a delinquent consumer

The Attunely propensity model estimates the likelihood of immediate payment for each delinquent account. Our model also produces an overall probability score that changes dynamically in response to the ongoing interactions between your organization and the consumer.

Liquidation Model

Analyze historical data to inform the expected value

The Attunely liquidation model combines behavioral signals with your organization's de‑identified historical transaction data to produce an expected value when liquidating debt. It is informed by historical calls, letters, e‑mails, and text messages, and dynamically refines an account-level score based on each interaction with the consumer.

Omnichannel Model

Identify the highest-yielding form of communication

The Attunely action impact scores rank each communication channel and identify the highest-yielding form of outreach on every account. Combined with the liquidation score, this strikes the optimal balance between immediate recovery, maximizing long-term recovery, and providing the consumer time to repay.

Time-of-Day Model

Match each consumer with a preferred time to be called

The Attunely time-of-day model recommends the best time to call each consumer to achieve success. Combined with our liquidation scoring model, we produce a dialer-ready call file that matches the highest-yielding accounts with their preferred time slots. This allows for more contacts using fewer call attempts.

Settlement Optimization Model

Calculate the best offer to settle an account in a debt purchase portfolio

The Attunely settlement model leverages de‑identified historical collection data to estimate the likelihood, timing, and expected liquidation value of accounts in a debt purchase portfolio. This enables recovery experts to produce the optimal settlement, payment plan, or paid-in-full demand for every purchased debt account.

Delinquency Prevention Model

Estimate risk and identify the best channel for outreach

The Attunely pre-delinquency model estimates the risk of an account entering delinquency. We leverage de-identified behavioral analytics and detection of anomalies to identify which accounts are most at-risk, recommend the most effective communication channels for preemptive outreach to prevent delinquency, and give consumers the opportunity to take advantage of programs designed to mitigate delinquency.

Default Prevention Model

Determine which consumers need counseling or additional support

The Attunely default model determines which accounts require minimal intervention and which would benefit from additional counseling or support to prevent default. This model provides a dedicated approach to handling alternative successful outcomes, such as deferrals, forbearance, or alternative options.

Agent Productivity Model

Evaluate agents to match them with specific accounts

The Attunely agent productivity model provides insights into the performance of each collection agent with each type of account. Building on our time-of-day technology, this model enables more effective staffing strategies and increased liquidation by matching individual agents with specific accounts and appropriately balancing their workloads.

Harris & Harris Improves Collections with Dynamic ML Scoring & Segmentation

Case Study

Credit Collection Partners Taps Attunely’s Machine Learning to Drive Increased Recoveries

Case Study