Executive Summary

The most widely-used database management systems (DBMSs) today are open source. Open source solutions such as MySQL and PostgreSQL have become ubiquitous in the enterprise, with more and more organizations embracing polyglot database environments (i.e. those that use multiple types of databases within the same application or ecosystem) as their standard system architecture.

Driven by an explosion in the volume and variety of data and data applications—and a “best tool for the job” mindset—these multi-database environments offer a litany of benefits and advantages across the modern enterprise, including: use-case-based optimization, greater agility and flexibility for developers, and improved system reliance and scalability.

To overcome fragmentation, data silos, and runaway financial and operational burdens, open, unified, AI-ready data infrastructure will soon be prerequisite for enterprise success. Moving forward, successful database management will be largely determined by decisions around servicing AI workloads, adopting cloud-native solutions, and increased scrutiny around licensing (even within the “open” software world).

These changing environments also present some unique challenges. With more to manage and more at stake, organizations that invest in tools, technologies, and services that help to mitigate these unique challenges stand to gain significant competitive

Here are some recent research findings from across the industry that help to frame and illustrate the state of open source database management in 2025:

Key Statistics

  • In a recent survey conducted by Percona, 78% of technical professionals said that PostgreSQL was important to their organizations’ current or planned AI/ML initiatives. With 25% saying PostgreSQL was “mission critical” to those initiatives.

  • According to the 2024 Data on Kubernetes (DoK) Community Report, databases are the #1 workload on Kubernetes for the third consecutive year. And nearly half of organizations now run 50% or more of their database workloads in production on Kubernetes, with the most advanced running over 75% in production.

  • In a recent survey, 70% of respondents with Redis deployments said the shift in Redis’ licensing had motivated them to seek alternatives, with nearly 76% saying they’ve either already migrated or are testing/ considering adopting Valkey.

Three Key Decisions for the Year Ahead:

  1. AI Readiness & Vector Strategy: The decision between investing in versatile, open source vs. proprietary, dedicated AI/vector DB solutions is a fundamental one for business leaders to make in the near term.
  2. License risk policy: How you will evaluate “open source‑ish” licenses and avoid single‑vendor projects.
  3. Control plane / IDP for data: Whether you’ll build or buy a unified, open, cloud‑native DB platform that abstracts ops, observability, security, and costs across increasingly diverse database environments.

Executive Playbooks: Navigating the New World of Database Management

Your 90‑day plan

• Inventory & score every DB you run with the Openness Scorecard + Unified Platform Scorecard (12 KPIs).

• Decide the default vector path (open source vs. proprietary, dedicated vector DB) and socialize the decision tree.

• Stand up a thin control plane pilot (or shortlist vendors) for provisioning, backups, observability, cost, and policy.

• Pilot AI copilots for DB ops

Your 12‑month plan

• Productionize the control plane (self‑service, SLO coverage, automated runbooks).

• Move to policy‑bounded automation for routine DB ops (patching, failover, scaling).

• Adopt FinOps for data: normalize unit costs (per TB / per query / per workload).

• Evolve from copilots  agentic ops with guardrails (audit trails, rollback plans).

• Continuously re‑score vendors against the Openness Scorecard to avoid creeping lock‑in.

Continue reading