On NEXTGRES and the evolution of PostgreSQL. Q&A with Jonah Harris.

Q1: What is NEXTGRES?

NEXTGRES is a next-generation, enterprise-class evolution of PostgreSQL designed to unify diverse data types, code execution, and computational workloads within a single, scalable system. By integrating multiple data models–including relational, document, time-series, vector, and others–NEXTGRES is dedicated to eliminating data silos and simplifying the data management stack for organizations. Other enhancements we are developing, such as advanced connection pooling, in-database code execution, and improved query instrumentation, are intended to reduce operational complexity, lower costs, and accelerate innovation.

Q2: What primary problem are you solving, and why is it a significant challenge for companies today?

The core issue we are addressing is the painful trade-off between developer productivity and long-term data value, which often presents as data fragmentation within enterprise environments. Over the years, we’ve seen organizations regularly prioritize quick development solutions, leading to the deployment of specialized databases to meet specific requirements like document storage, time-series data, or special-purpose search capabilities. While these systems solve immediate needs, they spread data across multiple platforms, creating silos that increase complexity and operational overhead.

This fragmentation makes it nearly impossible to effectively deliver enterprise-wide application development, analytics, and AI initiatives. And, companies often face higher costs, slower decision-making, and increased governance risks as well. To solve this, they’re typically forced to implement complex migrations, introduce data movement workarounds, or rely on expensive third-party virtualization technologies. By enhancing Postgres to support these diverse data models and workloads within a single platform, NEXTGRES unifies data management.

Our approach seeks to simplify the technology stack, reduce costs, and empower strategic decision-making, ensuring developers remain focused on innovation without being bogged down by the complexities of multiple systems.

Q3: Can you share a success story highlighting the impact of NEXTGRES?

Absolutely. A team we are working with was developing an AI project in the health & wellness space. Their infrastructure consisted of various MySQL and Postgres databases, as well as a number of microservices presenting data via web-based APIs. The application made heavy use of retrieval augmented generation (RAG), where its underlying data was stored in multiple database systems and required a significant amount of data pipeline work to combine, clean, and generate embeddings from, which were eventually stored in FAISS.

Given the number of different systems involved, the application had to manage and maintain a significant amount of configuration and database-specific data fetching logic. Our NEXTGRES POC consolidated a subset of that data, replaced the entire collection, cleansing, and embedding generation data pipeline with a single SQL query, allowed them to train and perform inference against a ranking model that optimized RAG results using in-database machine learning, provided transparent data virtualization to present microservice data in the same structure as the local data, and allowed them to perform RAG in a single SQL query. This simplified deployment and configuration, eliminated an entire data pipeline, colocated vector search into the same database as the data, and consolidated two databases into one, reducing operational management costs.

Compared to their similar projects, not only were the RAG results evaluated to be more relevant and precise, but the result was a 2x improvement in development time, approximately 4x faster application response time, around a 65% reduction in code written, and a roughly 22% reduction in project infrastructure costs.

By consolidating systems and leveraging in-database machine learning and vector search, the team not only streamlined their workflow but also achieved significant performance gains. This approach demonstrated the power of a unified data platform in simplifying complex AI pipelines, enabling faster iterations, and reducing both technical debt and operational overhead.

Q4: What makes it different from other Postgres distributions, like EDB?

There are many different Postgres variants out there. We believe NEXTGRES distinguishes itself by handling multiple specialized functions within a single platform. Unlike other variants that target specific use cases, we are transforming Postgres into a cohesive multi-model, multi-workload platform that natively integrates the key data management features enterprises expect, eliminating the need for multiple database systems and the overhead of managing complex data pipelines or migrations.

While other databases constrain developers by requiring them to learn entirely new technology stacks and build applications in specific ways, we’re leveraging the extensive PostgreSQL developer ecosystem and knowledge base. Our enhancements are designed to be additive, allowing developers to continue building as they always have – only now, their applications run faster on NEXTGRES. Some of the improvements we’re building in developer productivity and operational efficiency include advanced connection pooling for high-concurrency workloads, optimized in-database code execution to reduce latency, and enhanced query instrumentation for deeper performance insights. These are essential features that enterprises need but niche Postgres variants do not currently address. Additionally, our compatibility with various database APIs, including SQL Server, MySQL, and others, is intended to streamline migrations and infrastructure consolidation without requiring extensive application rewrites.

Ultimately, NEXTGRES is engineered to provide the comprehensive functionality of specialized databases combined with the simplicity, scalability, and cost-efficiency required to address modern data management challenges in analytics, AI, and application development.

Q5: You’ve mentioned that specialized databases are problematic. However, they only exist because traditional databases couldn’t handle specific use cases. What’s wrong with that?

You’re absolutely right; special-purpose databases emerged to fill gaps where traditional databases fell short, addressing needs like graph processing or handling time-series data. However, over time, established databases like Postgres have evolved to incorporate many of these specialized capabilities, such as supporting JSON for document data, native time-series functionalities, and vector search capabilities.

The problem with relying on multiple specialized systems is that it leads to data silos, limiting the value of data beyond the specific application it serves. Additionally, data integration becomes complex and error-prone when combining information from different systems for development, analytics, or AI. Not only that, but each additional database adds operational overhead in terms of maintenance, backups, security measures, and the need for specialized expertise. This not only increases costs but also kills a company’s ability to make timely, data-driven decisions.

We’re committed to eliminating the need for multiple specialized databases. This simplifies architecture, makes data more accessible and usable across the organization, and eliminates inefficiencies. We’re striving to fulfill the long-held dream of a stable, all-in-one platform that enables companies to focus on innovation rather than infrastructure management.

For the more technically inclined, an excellent paper on this topic is “What Goes Around Comes Around… And Around…” by database visionary and POSTGRES creator Michael Stonebraker and his well-known colleague Andy Pavlo..

Q6: What experiences or insights inspired the development of your solution, and how have these influenced your approach?

Over the past two decades, my career has been filled with observing and solving the same problems across various companies over and over again. I’ve seen countless organizations struggle through massive Oracle and SQL Server migration efforts, with timelines stretching into years and costs piling up into the millions, all because their data was locked in silos. Too often, their teams cobbled together brittle pipelines with a combination of open-source projects like Kafka and Flink, expensive pipelines with tools like Confluent, and even homegrown solutions, just to shuffle data between systems. These patchwork solutions almost never hold up over time or under pressure.

Time after time, executives asked me the same question: “Why can’t we just have one platform that works?” That frustration has been a significant inspiration for developing NEXTGRES. We’re building a solution that addresses these challenges step by step, delivering value at each stage. By incrementally enhancing Postgres to meet these needs, we’re providing practical solutions that offer immediate benefits while laying the groundwork for the future. Our approach is shaped by real-world challenges, striving to provide the simplicity and scalability enterprises actually need.

Q7: Who are the primary users or decision-makers at companies that benefit from your product, and what pain points does it specifically address for them?

The primary users of NEXTGRES are software developers, database administrators (DBAs), and data scientists, while the key decision-makers include CTOs and CIOs. NEXTGRES is designed to address several core pain points by unifying the data management platform, which simplifies the technology stack, eliminates complex pipelines, and reduces maintenance efforts.

For developers and DBAs, NEXTGRES offers a simplified development environment by supporting multiple data models and enabling consistent coding practices across applications. This reduction in complexity allows them to focus more on innovation and less on managing disparate systems, thereby enhancing productivity and streamlining maintenance tasks.

Data scientists and analysts benefit from consolidated data access, facilitating more effective analytics and the development of machine learning models. With unified and comprehensive data sets, and in-database machine learning, they can perform data preparation more efficiently, leading to faster and more accurate data-driven decision-making.

For CTOs and CIOs, NEXTGRES significantly lowers storage and operational costs by consolidating multiple database systems into a single platform. This consolidation not only reduces financial overhead but also enhances operational efficiency by eliminating the need for complex data pipelines and minimizing maintenance efforts. As a result, executive leaders can allocate resources toward strategic initiatives and innovation, rather than being bogged down by managing multiple disparate systems.

Overall, NEXTGRES enables organizations to reduce data management costs by 20-40%, simplify administration, and focus on innovation, providing substantial benefits tailored to the needs of each user group and decision-maker within the enterprise.

Q8: You noted database compatibility, which is your longtime area of expertise. With all the systems available today, how vital are interoperability and compatibility?

Interoperability and compatibility remain absolutely crucial, and they tie back into many of the challenges we’ve been discussing. While moving data between systems can be technically straightforward, the real challenges come from changing applications and ensuring everything works correctly after a migration. Incompatibilities can lead to significant downtime, higher training costs, and ongoing support issues. These challenges can cause companies to delay migrations or avoid them altogether, keeping them locked into suboptimal systems.

By ensuring high levels of compatibility, we allow organizations to migrate their databases without necessitating extensive code rewrites or risking application stability. This minimizes disruption during transitions, lowers the total cost of ownership, and accelerates the adoption of more advanced and scalable database solutions. Additionally, interoperability ensures that existing tools and workflows remain functional, preserving developer productivity and reducing the learning curve associated with adopting new technologies.

Q9: Is that why you’re making NEXTGRES compatible with MySQL applications? What are the benefits?

Exactly. Making NEXTGRES compatible with MySQL applications is a strategic decision that aligns with our overall goal of unifying data management and simplifying migrations. Many organizations are already moving their MySQL workloads to Postgres to take advantage of its advanced features and robust ecosystem. However, migrating applications can be challenging due to differences in SQL dialects, data types, and behavior.

By extending compatibility, NEXTGRES enables these organizations to migrate MySQL applications with minimal code changes, reducing both time and cost. After migration, they can leverage NEXTGRES’s advanced capabilities, such as multiple data models, in-database machine learning, and more robust extension support, which are not available in MySQL. This compatibility reduces the need for code changes, speeds up adoption, and allows organizations to consolidate their systems while taking advantage of advanced Postgres features and thriving ecosystem.

Consolidating systems reduces operational overhead and complexity, lowering infrastructure costs and simplifying administration. Ultimately, this compatibility provides a practical pathway for organizations to modernize their database infrastructure while preserving existing investments in MySQL-based applications, furthering our mission to provide a unified, scalable platform.

Q10: Is there anything else you wish to add?

I’d only emphasize that none of the challenges enterprises face with data management are new, and the solutions thus far have all been piecemeal – focused on short-term fixes rather than long-term value. NEXTGRES represents the culmination of years of observing these issues and understanding what organizations genuinely need: a unified platform that integrates seamlessly into existing systems while providing the scalability and features necessary to remain competitive.

By addressing this in one system, we’re not just creating a better database; we’re enabling companies to unlock the full potential of their data without sacrificing developer productivity. My team and I believe the future of enterprise data lies in this unified approach, and we’re committed to delivering a solution that evolves with the needs of modern businesses.

……………………………………………..

Jonah H. Harris, Founder of NEXTGRES

Jonah is a seasoned developer and entrepreneur with a deep focus on database internals, high-performance systems, compatibility methods, network protocols, and near-real-time machine learning optimizations. He excels in data structures, algorithms, and software optimization, establishing himself as a recognized expert in Oracle Database and a longstanding contributor to PostgreSQL. Throughout his career, Jonah has held pivotal leadership roles, including Chief Technology Officer at MariaDB (NYSE) and The Meet Group (NASDAQ), and he was a founding engineer at EnterpriseDB. Currently, Jonah is the co-founder of NEXTGRES, where he continues to drive innovation and excellence in the database technology landscape.

Sponsored by NEXTGRES