Frequently Asked Questions

Payment leakage. Claims are paid incorrectly because policies are stored in PDFs, contracts drift, and clinical complexity masks errors. We turn those documents into executable rules so payers can detect, prevent, and defend every dollar.

Legacy tools store documents; we compute them. Our AI-native, neuro-symbolic engine extracts facts, verifies them against rules, and supports decision-making with clause-level provenance.

Claims Repricer, Policy Intelligence, Contract Library, and DRG Review & Reclassification—all feeding Payment Leakage Analysis and predictive ranking.

No—nēdl augments coders, nurses, clinicians, SIU/FWA, and existing vendors. Most customers run nēdl alongside current workflows to expand coverage and reduce rework. We can review 100% of claims, policies, and contracts at scale and speed.

We transform policy and contract clauses into typed, temporal, and code-set-aware rules that can be executed on every claim, with complete lineage back to the source text.

Reprice (e.g., Medicare relativity). 2) Layer policies & contracts to surface drift/discrepancies. 3) DRG review parses clinical notes and history for upcoding/bundling. 4) Evidence pack with citations & version diffs routes to post-pay recovery or pre-pay edit.

Every extracted fact is validated by rules (typed/temporal/code-set constraints) and cross-checked in a knowledge graph (e.g., REQUIRES/EXCLUDES/OVERRIDES), producing explainable flags.

Yes, our pipeline is built for continuous, risk-based monitoring of all claims, not small samples. High-risk items are prioritized for human review.

Customers use nēdl to detect more leakage, resolve faster, and lower provider abrasion with evidence-backed decisions. Modeled programs target double-digit ROI and measurable PMPM improvement.

We tie detections to recoveries or prevented dollars, plus operational savings from faster, fewer appeals. Rule-of-thumb PMPM: PMPM = Dollars / (Members × 12) (e.g., ~$12M per 1M members ≈ $1 PMPM).

Most start post-pay to harvest near-term value and proof, then move left to pre-pay using the same validated rules to prevent errors before payment.

Claims (X12 837/UB-04/HCFA), remits (835), policies (CMS & payer), provider contracts/addenda, clinical notes/medical history when applicable, and optional reference files (fee schedules, code sets).

Options include secure REST APIs, SFTP drops, lakehouse connectors (e.g., Snowflake/Databricks), or batch files. We produce result feeds, evidence packs (PDF/JSON), and dashboard views.

Typical pilot launches take 4–8 weeks, depending on data readiness and scope (lines of business, modules enabled).

A compound approach: AI extracts and summarizes, symbolic rules verify facts deterministically, and a knowledge graph encodes relationships (REQUIRES/EXCLUDES/OVERRIDES/SUPERSEDES) for multi-hop reasoning.

Every decision carries clause-level citations, timestamps, version diffs, and data lineage—so auditors and providers can see precisely why a claim was flagged or repriced.

Yes—Graph-RAG answers multi-step queries (e.g., "Show all TKA policies with BMI < 40 and effective dates"), returning citations to the governing clauses.

We design for HIPAA and sign BAAs. Data are encrypted in transit and at rest, with least-privilege access, audit logging, and customer-defined retention.

We deploy on Microsoft Azure with options for single-tenant environments, private networking/VNet peering, customer-managed keys (e.g., Azure Key Vault), and regional data residency.

We support tokenization and pseudonymization, as well as scoped datasets, for the contracted use case and duration.

A data contact, a workflow owner, and secure data pipes. We provide mapping templates, rule authoring sessions with your SMEs, and pilot playbooks.

Continuous ingest + compare detects deltas; impacted rules are re-verified and promoted with change logs. You get alerts and a one-click diff view.

Flags flow to your UM/PI queues; evidence packs shorten appeals and lower overturns. SIU/FWA teams can subscribe to high-likelihood patterns.

Two components: SaaS (platform/modules) plus an optional outcomes-based share for validated recoveries/prevented dollars. Pilots are scoped with clear success metrics.

Yes. We power partner engage with policy-as-code and evidence packs, while partners provide clinical operations and payer relationships. We can become an AI layer for vendor solutions, augmenting their people and tech to achieve more quickly and at scale.

The opposite is the goal. Clause-level evidence and provenance reduce back-and-forth and improve acceptance, lowering disputes and overturns.

Yes—decision summaries include the governing clause, effective dates, thresholds/codes, and how they applied to the claim.

A 60–90-day program with defined lines of business, target edit sets, success metrics (e.g., $ recovered/prevented, TAT, abrasion), and weekly checkpoints.

Shared dashboards track detection quality, recovery/prevention, operational savings, and PMPM impact, with independent finance sign-off where required.

Dollars paid incorrectly due to coding, policy/contract drift, eligibility, duplicates, or DRG issues.

When adjudication deviates from the currently effective policy or contract clause.

Repricing commercial claims using Medicare as a reference anchor.

Inpatient prospective payment system; DRGs group inpatient stays for payment and can be upcoded or bundled.