“We’re building the AI infrastructure healthcare doesn’t yet realize it needs.”

Tashkent, Uzbekistan (UzDaily.com) — As the U.S. healthcare system faces increasing administrative costs and claim denials, data-driven innovation has become a strategic priority. Umidjon Saidkhujaev, an Uzbek-born expert in business intelligence and AI technologies, is among the leaders shaping this field. Drawing on experience from transformation projects in Uzbekistan, analytics roles at Axtria in the United States, and a research-focused MBA from Lehigh University, Saidkhujaev is now launching ClarityClaim AI, an early-stage U.S. startup focused on vertical AI for claims denials and appeals.

In this interview with UzDaily, Umidjon Saidkhujaev discusses his career, the gaps he identified in U.S. healthcare, and why he believes now is the right time to develop AI infrastructure for this multibillion-dollar challenge.

— You’ve worked across Deloitte, NMMC, Asaka Bank, Axtria, and now your own startup. How did your focus shift toward healthcare?

— My focus shifted during my MBA at Lehigh University, where I examined the U.S. healthcare system from data, policy, and operations perspectives. Lehigh’s research-driven environment highlighted the system’s fragmentation and administrative complexity, as well as the potential for analytics to drive improvement.

My role at Axtria accelerated this transition. Axtria is central to U.S. pharmaceutical analytics, and my work there exposed me to a wide range of healthcare data, including claims feeds, prescriber behavior, payer rules, model monitoring, and compliance reporting. Managing billions of dollars in healthcare data each quarter revealed both the system’s strengths and its outdated administrative processes.

— What did you learn at Axtria that pushed you toward building your own platform?

— Two lessons stood out.

First, healthcare data engineering differs significantly from other sectors. It requires interpreting payer policies, understanding regulatory constraints, reconciling medical codes, and ensuring predictive models perform reliably with complex real-world data. At Axtria, we built analytical engines that pharmaceutical companies relied on for critical revenue decisions, making accuracy essential.

Second, I observed a structural disconnect. While pharmaceutical companies invest in advanced machine learning for targeting and forecasting, hospitals often rely on manual processes, such as PDFs, spreadsheets, and copy-paste letters, when handling claim denials and appeals. This imbalance highlights the need for an AI infrastructure layer in healthcare, particularly in claims denials and appeals, where cost, equity, and access are closely linked.

— ClarityClaim AI is still in a very early stage. Why build the company now?

— The technology now makes this possible. Until recently, expecting AI to read payer manuals, generate appeal letters with citations, or identify inequities in denial patterns was unrealistic. Today, advanced language models, retrieval-augmented generation, and mature MLOps practices enable us to design AI systems that are transparent, auditable, and compliant with healthcare regulations.

This is the moment when the idea, the technical architecture, and the market need to align.

— You’ve been conducting market research with hospitals. What trends are most urgent?

— Healthcare providers consistently highlight three growing pressures: denials are increasing faster than staffing can manage; equity concerns are rising as some demographic groups face higher denial rates; and regulatory expectations are tightening, with agencies like CMS requiring stronger documentation of decision logic.

Revenue-cycle leaders often say, “We don’t need another dashboard. We need a system that can read, interpret, explain, and act.” This feedback is why ClarityClaim is built on vertical AI, a system tailored to a specific domain rather than a general-purpose model.

— How do you explain “vertical AI” to executives who are exposed to endless buzzwords?

— I keep it simple:

Horizontal AI is ChatGPT; vertical AI is what a compliance team trusts during an audit.

Vertical AI specializes in a single domain. For us, it understands claims denials and appeals, including Medicare and Medicaid criteria, payer rules, prior authorization logic, and medical coding standards. Most importantly, it explains its reasoning with citations, which is essential in healthcare.

— What aspects of ClarityClaim’s technology do you consider non-negotiable?

— Two components define our approach: the policy knowledge graph, a structured representation of healthcare rules, and the explainability layer, which ensures that every AI-generated recommendation includes its source, rationale, and supporting evidence.

Explainability must be a priority. In administrative workflows, especially appeals, trust relies on clarity. Equity analytics is also essential. If the system identifies that certain populations face disproportionate denials, it must alert providers rather than reinforce the pattern.

— People often assume “automating appeals” means simply generating letters. How do you address this misconception?

— Automating appeals is not just about generating text; it is about automating reasoning. This involves identifying root causes, interpreting payer logic, validating coding structures, and assembling evidence. Letter generation is simply the final step in a comprehensive analytical process.

— What does success look like for ClarityClaim over the next 18 months?

— Our goals are clear and foundational: finalize the technical architecture, complete comprehensive market research, secure early design partners, release an MVP for denial prediction and policy-backed appeal drafting, and implement a governance framework aligned with the NIST (National Institute of Standards and Technology) AI Risk Management Framework.

Scaling prematurely is not our priority. Establishing trust through accuracy, transparency, and speed is our main focus.

— You’ve built systems for governments, banks, pharma, and now healthcare. What connects these experiences?

— Each of these sectors involves large-scale, regulated environments where decisions have significant financial and human consequences. Across mining, banking, pharmaceuticals, and healthcare, the core challenge remains the same:
How can we make decisions faster, more accurate, and more fair?

I have addressed this question in various contexts throughout my career. ClarityClaim AI represents the most urgent domain where this expertise can have a meaningful impact.

— Why U.S. healthcare, and why now?

Saidkhujaev: No other sector presents such high stakes for patients, providers, and the national economy. Administrative complexity is undermining access, equity, and financial stability throughout the system. For the first time, technology enables us to build AI systems that bring transparency and fairness to these processes.

Behind every denied claim is a hospital, a clinician, and a patient navigating a system that has become too complex to manage manually. If AI can help reduce that burden, even slightly, it is worth building.