Zero Budget for Error: Safeguarding Terabytes of Medical Data In this interview, Nithinreddy Burgula discusses how to ensure the seamless operation of databases in mission-critical projects, manage terabytes of confidential data with precision, and highlights the essential role well-coordinated technical teams play in these complex processes.

The healthcare sector generates more than 30 per cent of the world's total data — a share expected to increase to 36 per cent by the end of 2025. In developed countries, the average hospital produces approximately 50 terabytes of data annually, including patient health records, test results, medication prescriptions, and other essential information. This volume continues to grow as treatment protocols advance, healthcare regulations evolve, and new drugs and diagnostic methods are introduced.

A portion of this data is shared with insurance companies, which use information on doctor prescriptions, patient health status, and medical history to calculate premiums and assess payout risks.

Data quality, integrity, and security are fundamental to success in the insurance sector — issues that Nithinreddy Burgula, an experienced database administration expert, has been addressing for years. With six years of experience ensuring the security, reliability, and accessibility of databases at major U.S. corporations — including industry leaders like Chevron and ESTI — he has spent the last two years focused on maintaining the stability, performance, and resilience of databases at Cigna Healthcare, one of the largest and most influential companies in the American insurance market.

In this interview, Nithinreddy Burgula discusses how to ensure the seamless operation of databases in mission-critical projects, manage terabytes of confidential data with precision, and highlights the essential role well-coordinated technical teams play in these complex processes.

From Neural Network Theory to Big Data Management

Nithinreddy Burgula's interest in database administration began during his master's studies in Computer Science at Lewis University. His coursework highlighted the management of large-scale data systems for enterprise-level projects and the training of artificial intelligence models. Several academic projects also involved MongoDB, a NoSQL database management system.

By the end of his studies, databases had become his primary professional focus. After earning his master's degree, he decided to deepen his expertise in administration and completed a one-year program at MongoDB University — a training school established by the creators of the database management system.

Nithinreddy's career began with an internship at Chevron, one of the largest energy corporations in the U.S. As a junior database administrator, he worked on projects aimed at improving database performance, implementing backup solutions, and managing the migration of specific replica sets and data clusters.

His experience collaborating with cross-functional teams — from Ops engineers to developers seeking DBA support for assessing software updates, mitigating security risks, ensuring database stability, or optimizing query response times — became a key asset. This practical expertise helped Nithinreddy secure a contract with ESRI, a company specializing in geographic information systems.

"At ESRI, data is the backbone of the company's products. They needed to ensure database stability, enhance data security, and create a reliable encryption process. To help streamline operations and free up time for the main tasks, I automated routine processes. Using Python scripts, I was able to handle backups, monitoring, and index updates on a regular schedule, which not only saved time but also minimized the risk of human error. This gave me the opportunity to focus on more complex and high-priority challenges, where my skills were put to even better use," explains Nithinreddy.

After his time at ESRI, Nithinreddy worked at UST, a global provider of business transformation technology solutions. In 2023, he joined Cigna Healthcare, one of the largest companies in the U.S. health insurance sector.

How Insurance Companies Are Tackling the Increasing Data Demands

By the time Nithinreddy joined Cigna Healthcare, 92.5 per cent of the U.S. population had health insurance, and each major player in the industry was managing petabytes of customer data. This volume was continually increasing, as individuals sought medical care, received doctor's prescriptions, test results, and diagnoses, and updated personal information such as addresses and names.

Insurance companies receive customer data in a variety of formats — including PDFs, printed copies of medical reports and consent forms, text contracts, and images. All of this data must be processed and standardized. A single incorrect entry or mistake during data cleaning can lead to errors and misinformed decisions. For example, if an insurance company lacks sufficient data on a policyholder, it could mistakenly deny coverage.

The consistency and stability of a database directly impact success in the insurance industry. When pricing policies, insurers need to factor in all risks — such as medical history, age, and past claims — since these elements affect both profitability and customer satisfaction.

Moreover, data must be securely stored in compliance with the Health Insurance Portability and Accountability Act (HIPAA). Medical information about policyholders is confidential, and any data breach could lead to severe legal and financial consequences, not to mention a loss of patient trust.

One of the first major challenges Reddi faced at Cigna was migrating a large, high-load replica set to new hardware using MongoDB — all without causing any downtime. This wasn't a simple data transfer; the project required a deep understanding of MongoDB's replication mechanics. Careful planning was essential: Reddi first deployed a parallel replica set on the new infrastructure, then added it as a secondary node to the existing production set. Continuous monitoring of replication lag was critical to maintaining data consistency.

The final stage involved a seamless primary node switch: one of the new secondary servers was promoted to primary, and the old nodes were carefully decommissioned from the cluster. This project gave Reddi invaluable hands-on experience with live system migrations, configuring fault-tolerant setups, and minimizing production risks.

Addressing the Practical Challenges of Data Management

Nithinreddy's team at Cigna Healthcare is responsible for ensuring data compliance. As a senior database administrator, he oversees data security, encryption, authentication, certificate management, and regular data backups. He also ensures that data is recorded and stored correctly, and that it's always accessible.

"We handle medical data, so making sure everything is available 100% of the time—both during business hours and after—is absolutely essential. For instance, when we're updating our database or internal systems, we know in advance that some services will be temporarily down. We have to give our clients and partner clinics a heads-up at least a week in advance. That means we can't afford any mistakes—our system has to run smoothly and stay stable, even when there are failures. Getting to this level of reliability, especially with the sheer amount of data we manage and optimize, has been one of the toughest challenges of my career," explains Nithinreddy.

Along with his team, Nithinreddy focuses on data replication, works on reducing latency in database queries, ensures the availability of all systems, and regularly audits database performance. He also automates routine database maintenance tasks to boost operational efficiency.

"Another key part of the job is keeping data consistent and making sure we back it up regularly. That way, if anything goes wrong, we can restore the data quickly without losing a single byte," adds Nithinreddy.

In one project, Reddi's team faced a critical issue with low database throughput. Data loaded painfully slowly, users complained about long wait times, and the system struggled to handle the load.

"It felt like a highway jammed with massive traffic," Reddi explained. "In this project, I was searching for bottlenecks — operations that were too slow and dragging down the entire system. Our investigation revealed multiple issues: some queries were inefficient, so we added indexes to optimize them. Sticking with the traffic analogy, it was like building additional fast lanes to boost the highway's capacity."

Collaborative Strategies for Preventing Database Failures

To ensure the service runs smoothly, effective communication between all teams involved in development is crucial. For example, if Ops managers are planning a release that might affect database performance (like suddenly increasing the number of server requests), the DBA team needs to be in the loop. This becomes even more important when the company owns multiple products that rely on a shared database. For instance, in addition to its core service, Cigna Healthcare also manages Express Scripts — an app that handles the distribution and pricing of discounted medications at pharmacies.

Once a month, the team conducts failure drills to ensure the system can quickly recover and that no data is lost. In addition, Nithinreddy sets up logging and monitoring systems for user activity, which track all events happening with the database in a special log. This log allows them to see who accessed the database, when, and with what level of permission — plus, whether the action was authorized.

"My team and I regularly update access and encryption keys. The more frequently passwords or keys change, the harder it is for someone to intercept and use them to gain unauthorized access to the database," adds Nithinreddy.

How to prevent leaks in critical infrastructure

Another critically important aspect of working with medical data is security and protection against breaches. One of Reddi's key tasks at Cigna was to implement robust encryption without compromising database performance.

After weeks of research, Reddi's team integrated the KMIP (Key Management Interoperability Protocol) encryption standard, achieving an optimal balance between performance and data security.

"Implementing the solution required deploying a dedicated key management server (KMS), integrating it with MongoDB Enterprise Advanced, and configuring encryption for collections containing protected health information (PHI)," Reddi explained. "The key server was set up with a strict key rotation policy and tightly controlled access management, ensuring secure storage and regular updates of encryption keys. KMIP enabled centralized key management and seamless compatibility with other security systems."

Beyond disk-level encryption, Reddi also helped design the data transmission system and configured mandatory encryption for all connections to the MongoDB cluster using TLS/SSL. This included setting up MongoDB servers to accept only encrypted connections, distributing certificates to client applications, and ensuring end-to-end encryption of traffic between application servers and the database, as well as between replication nodes.

From Database Management to AI

The volume of data in the medical insurance industry is growing at an incredibly fast pace. In response, market participants are gradually beginning to implement artificial intelligence (AI) and machine learning algorithms to automatically process these vast amounts of data.

Manual analysis is no longer sufficient for assessing insurance cases due to the sheer volume of data. Human evaluation is time-consuming, and technology offers a more efficient way to assess risks with greater accuracy, enhance client retention, and improve overall satisfaction.

Artificial intelligence development and data security are the key areas where Nithinreddy plans to grow his skills, leveraging his extensive hands-on experience in database administration.

"My goal for the next few years is to focus on building sophisticated and effective data protection solutions," Reddi shared. "This includes advanced encryption methods like homomorphic encryption and differential privacy, designing finely tuned access control mechanisms, developing data lineage tracking systems, and ensuring compliance with regulations such as GDPR, CCPA, and HIPAA. I'm especially excited to work in this space given the massive data volumes required for machine learning and AI development — it's a dynamic challenge that pushes the boundaries of security innovation."