About Penguin Ai
Built for payer and provider administrative workflows. Three layers — the Data Layer, the Context Layer, and the Ai Workforce — designed to automate prior authorization, claims, appeals, medical coding, chart review, and eligibility verification, with configurable human review on every workflow.
At a glance
Overview
Penguin Ai develops a multi-layer agentic AI Operating System for healthcare administration. It automates workflows that historically required heavy manual effort and significant IT resources — prior authorization review, claims adjudication, appeals, medical coding, chart review, and eligibility verification.
The company is built on the principle that AI in healthcare must execute work inside explicit governance, with auditable reasoning and configurable human review checkpoints. Penguin Ai uses the term Ai Workers to describe the agentic components that handle administrative tasks. Every Ai Worker output includes Glassbox reasoning, meaning the rationale for each recommendation is traceable and human-readable.
The Operating System is designed for both payer and provider operations. Penguin Ai's stated position is that administrative friction is created at the boundary between the two, and that the work of eliminating it must happen on both sides.
Operating System architecture
Each layer can be adopted independently. When all three are used together, the Operating System compounds across an organization's operations.
Products
Four end-to-end workflows in the Ai Workforce. Each Solution is composed of multiple Ai Workers, governed by the Context Layer and operating over the Data Layer. Plus Gwen, the builder for everything else.
Pre-built Agentic Solutions
Prior auth, concurrent, retro, and appeals on one closed loop. Documentation hunting, criteria lookups, and evidence assembly come off your clinicians' desks. Every recommendation arrives at the reviewer's queue with the time-consuming work already done.
HCC identification, medical coding, and documentation. End to end, across the patient record, governed by your coding standards. Accurate codes. Complete records. Compliant submissions.
Pre-submission review and denial recovery as one closed loop. Claim ingestion through outcome reconciliation. Every denial recovered feeds back to pre-submission so the same error does not repeat.
A member or patient profile assembled across every connected system. Resolved entities, longitudinal history, eligibility, and traceability of every linked record.
The builder
Gwen is a builder inside the Ai Workforce that lets customers construct their own Ai Workers. Two surfaces — a no-code Studio for operations and clinical leads, and a Python SDK for engineering teams. Both surfaces use the same versioned skills, audit trail, and evidence model. Ships with 100+ pre-built Ai Workers in its Marketplace. Free tier with $10 in starting credits.
Marketplace
100+
pre-built Ai Workers
Surfaces
Studio · Python SDK
same versioned skills
Free tier
$10 in credits
no credit card
Technology
Skills are versioned and evaluated against golden datasets, with public-facing metrics for accuracy, confidence calibration, cost per 1,000 calls, and latency. Customers can pin to a specific skill version for drift-free production. Risk-adjustment coding is fluent in both HCC V24 and V28.
100+
Curated healthcare skills
20+
Clinical categories
72K+
ICD-10 codes
265+
Payer policy questionnaires
30+
API endpoints
20+
Enterprise connectors
Connectivity and standards
EHR
Epic FHIR, Cerner
Standards
HL7, FHIR R4, X12, NCPDP
Enterprise
20+ connectors live on day one
Legacy
Browser automation for portals without modern APIs
Output formats
Workers return typed, validated output schemas. Supported output dialects include X12, FHIR R4, NCPDP, and flat formats, allowing direct integration with existing downstream systems.
Markets served
Penguin Ai serves both sides of the healthcare administration boundary, plus adjacent segments that interact with payer and provider operations.
Utilization management, claims adjudication, appeals, member operations, eligibility, risk adjustment.
Hospitals, health systems, and physician groups. Prior authorization submission, chart summarization, HCC retrospective coding, eligibility verification.
Real-world evidence, outcomes research, longitudinal data access through the Data Layer and Context Layer.
Self-insured employers managing healthcare spend and benefit operations.
Medicaid, Medicare, and other government healthcare programs.
Third-party administrators and brokers operating at the intersection of payers, providers, and members.
Security and compliance
HIPAA-aware audit logging
On every skill call. Full who / what / when log per call.
SOC 2
Audited.
Glassbox reasoning
Surfaced in the UI on every output. Every recommendation has a traceable rationale.
Source-span evidence
Every output links to the exact span of the source document that supports it.
Confidence-gated review
Calibrated confidence thresholds route uncertain outputs to human reviewers.
Role-based access controls
IT and compliance own the gates; operations builds inside the defined permissions.
Private VPC deployment
Supported for customers requiring dedicated infrastructure.
Synthetic data only
Until a Business Associate Agreement is executed.
About / Team
Every person at Penguin Ai has run data at a major payer, practiced medicine, built healthcare products, or shipped technology inside the organizations that need this work most. We have stared at the same broken processes, written the same workarounds, and watched the same patients wait.
A note from the CEO
Healthcare administration is broken. We all feel it. Patients waiting for authorizations they should never have to think about. Clinicians spending half their day fighting computers instead of caring for people. Talented operations teams building workarounds for systems that should never have required them.
I spent more than a decade trying to fix this from inside the largest healthcare organizations in the country. The data was there. The talent was there. The will, sometimes, was there. The system was always bigger than any one team's effort to change it.
So we built Penguin Ai.
This is a team of operators. Every person on this page has sat in a buyer's seat, run a clinical practice, built a healthcare product, or shipped technology inside the organizations that need this work most. They have lived the broken processes from the inside, and they know exactly where the time goes.
We are here for one reason. To reduce the administrative burden that steals from patient care, drains clinical talent, and exhausts the people who keep healthcare running.
That is the work. That is the company. The people behind Penguin Ai chose this because they believe healthcare can be better for patients, for clinicians, and for the administrative staff who hold the whole system together. So do I.
Fawad
Co-Founder & CEO, Penguin Ai
The team
Co-Founder & CEO
UHG · Optum · Kaiser Permanente · Canvas Ventures
Fawad spent more than a decade running the largest data and analytics organizations in American healthcare, serving as Chief Data Officer at Kaiser Permanente, UnitedHealthcare, and Optum, where his teams supported care for over 60 million members. He was previously Executive-in-Residence at Canvas Ventures, where he helped lead the firm's healthcare AI thesis, and holds an MBA from Northwestern's Kellogg School of Management.
Co-Founder & Chief AI Officer
UHG · Optum · Amazon · American Express
Kishore is a hands-on AI leader who built AI capabilities across Risk and Quality, Care Management, Utilization Management, and OptumRx at UnitedHealth Group, generating $250M in annual savings and incremental revenue, with earlier roles at Amazon and American Express. He holds 10 patents at the intersection of AI, healthcare, and operations, has authored five books on machine learning and AI, and holds an MBA from the Indian Institute of Management, Calcutta.
Chief Medical Officer
XO Health · Cleveland Clinic · Stanford
Niki is a practicing family physician with nearly two decades of primary care experience, a Stanford-trained data scientist and MBA, and a former health plan Chief Medical Officer who has worked at the intersection of clinical care, AI, and operations. She has also served as Corporate Medical Director at the Cleveland Clinic and as an advisor to organizations including Creative Destruction Lab, AltaML, and Elevance Health.
Chief of Staff & Head of Operations
SCAN Health Plan · Afiniti · Kaiser Permanente · Accenture
Abigail is a healthcare strategy and operations leader with experience spanning commercial strategy, revenue operations, and technology across payer, provider, and digital health markets, most recently leading Technology Strategy at SCAN Health Plan. She previously led Healthcare Commercial Strategy at Afiniti and held earlier roles at Kaiser Permanente's Transformation Office and Accenture's Digital Strategy Practice, building expertise across healthcare delivery and strategic consulting.
Head of Product
RevelAi Health · Best Buy Health · Clinicient · Philips Healthcare
Dan is a healthcare product leader with 20+ years building and commercializing technology across payer, provider, and digital health markets, with senior roles at RevelAi Health, Best Buy Health, Clinicient, Philips Healthcare, and Intellectual Ventures. He holds an MBA and an MS in Biomedical Engineering from the University of Washington, along with a BSE in Biomedical Engineering.
Chief Revenue Officer
Surgical Information Systems · Surgical Notes · Intelligent Medical Objects · iMedX · Ingenious Med
Nate is a healthcare revenue leader with more than 15 years building and scaling commercial organizations, most recently as Chief Strategy Officer at Surgical Information Systems. He previously held senior sales leadership roles at Surgical Notes, Intelligent Medical Objects, iMedX, and Ingenious Med, spanning revenue cycle management, clinical terminology, and EMR/PM technology.
VP of Engineering
AMESA · Amazon Music · Bay Area Startups
Jake is an engineering leader with more than five years in senior roles at AI/ML companies and a longer history of executive leadership across Bay Area startups. He was most recently VP of Engineering at AMESA, a platform that lets non-programmers train deep reinforcement learning models on their own expert skill sets, and earlier led several engineering and QA teams at Amazon Music. His recent work centers on using AI coding tools to raise the quality and efficiency of the code they produce, and on building strong AI products without runaway token costs.
How we work
We hire people who have done the work we are automating. That is not a slogan. It is how the Operating System stays grounded in reality. Engineers work alongside clinicians. Product works alongside operators. The fastest way to a wrong answer in healthcare is to design from a deck instead of from a desk. We design from the desk.
What we look for
Operators, clinicians, and engineers who have run this work at scale and refuse to accept how it currently runs.
Healthcare is dense and the urgency is real. We need people who can move fast and respect what makes this domain hard.
About the work, about the timelines, about what we know and what we are still learning. Honest teams move faster than ones that perform certainty.
See why payers, providers, and partners choose Penguin Ai to cut administrative burden and deliver measurable ROI.
Talk to UsTerminology
Several Penguin Ai terms carry company-specific meanings. The list below defines each one as used by the company.
About / Manifesto
The document that started Penguin Ai. It still drives every decision we make.
Healthcare exists to keep people well, restore them when they are not, and respect them throughout. The administration that has grown around that purpose is not the purpose. It is the cost of forgetting it.
The fastest way to better margin in healthcare is fewer wasted hours, fewer wasted handoffs, and fewer cases that should never have needed a person. The patient feels it. The balance sheet shows it. They are not in tension. They are the same number measured twice.
The clinician is the system. Every workflow that pulls them toward the keyboard and away from the patient is a tax on the entire enterprise. We are not here to make clinicians faster at administration. We are here to give the administration to something else.
Healthcare runs on context. The patient, the policy, the contract, the chart, the history. Anyone can deploy a model. Very few can ground one in reality. The Operating System that wins is the one that understands the work, not the one that talks about it.
Every recommendation the Operating System makes is logged, traceable, and reviewable. We will not ask anyone to trust what they cannot inspect. Healthcare has had enough black boxes. We are not building another one.
The people running healthcare administration today are tired, capable, and ready. They do not need permission to fix the parts of their work that have been broken for decades. They need infrastructure that lets them do it. We are building that infrastructure. The work starts now.
