Why Database Administrators Were "Going Extinct"
A recurring history of predictions and why DBAs are still here!
In the late 1990s, I began working as a paralegal for a small law firm specializing in business litigation. As my role developed and technology advanced, I also became a part-time database administrator (DBA), by taking care of our digital litigation databases. The more technology advanced, the more my role moved to a litigation support specialist with specific DBA-type duties of maintaining databases and preparing for in-court presentations. In the early 2000’s I made the jump to a full-time DBA with an IT consulting company. It felt like the next day I was reading how the DBA role would be phased out as it was no longer needed. Self-tuning databases were here and my job was going away.
More than 25 years later, I still work as a DBA on a DBA team. We support clients with on-premises databases and cloud-based databases. Many functions of our job have remained the same, but the lingo has changed. In the end, databases still hold data and have users. Performance is an issue. Quality DBAs are still needed. We must be adaptable as technology changes. Today, AI is not taking our jobs, but we must adapt and learn how to use AI. As Mark Twain said, “The report of my death was an exaggeration.”
Despite decades of predictions that new technology would eliminate the need for database administrators, the role has persisted — and evolved. Below is a rundown of the major waves of "DBAs are doomed" thinking, and why each one didn’t pan out.
Self-Tuning / Autonomic Databases (2000s)
Oracle, IBM, and Microsoft rolled out self-tuning features (automatic memory management, auto-indexing, query optimizers) with marketing that implied manual tuning would become unnecessary.
Reality
Tuning is only one slice of the job — capacity planning, security, architecture, and troubleshooting still needed humans.
NoSQL Movement (late 2000s–2010s)
"Schema-less" databases like MongoDB were pitched as freeing developers from rigid relational structures and, by extension, from DBAs.
Reality
NoSQL systems still need people managing sharding, consistency tradeoffs, and performance — the job just shifted; it didn’t disappear.
Cloud Databases / DBaaS (2010s)
Amazon RDS, Google Cloud SQL, and Azure SQL Database were marketed as eliminating the need for someone to "manage the database" since the cloud provider handled patching, backups, and provisioning.
Reality
Someone still has to design schemas, write efficient queries, manage costs, handle migrations, and secure data — the role moved up the stack rather than vanishing.
Big Data / Hadoop Era (early 2010s)
The hype suggested traditional relational databases (and their administrators) were being replaced wholesale by distributed data platforms.
Reality
Relational systems remained the backbone of most transactional workloads; Hadoop solved different problems.
DevOps and Infrastructure-as-Code (2010s)
As DevOps automated infrastructure provisioning, some assumed database administration would be automated away too, folded entirely into CI/CD pipelines.
Reality
DBAs adapted into "DataOps" or database reliability engineer (DBRE) roles instead of disappearing.
Serverless Databases (late 2010s–2020s)
Products like Aurora Serverless and DynamoDB were framed as needing zero administration — just write code, no database management.
Reality
Schema design, query optimization, and data governance are still required; "serverless" mostly hides infrastructure, not data modeling decisions.
Low-Code/No-Code Platforms (2020s)
Platforms promising that "anyone can build an app without knowing databases" suggested DBA skills would become irrelevant for most business use cases.
Reality
These platforms often hit scaling and data-integrity walls that still require expert intervention.
Autonomous/AI-Driven Databases (late 2010s–2020s)
Oracle’s "Autonomous Database" was explicitly marketed with taglines about eliminating human database administration through AI-driven tuning, patching, and security.
Reality
Adoption required significant oversight, and most enterprises kept DBA teams for exceptions, compliance, and hybrid environments.
Generative AI / LLMs Writing SQL (2020s, ongoing)
The current wave: AI copilots that write queries, suggest schemas, and even debug performance issues have prompted fresh "DBAs are obsolete" takes.
The DBA team at Moser Consulting is a group of high quality, highly qualified DBAs. We are adaptable. We do not run from change, we embrace it and work with it. Data, databases and users are not going away. In fact, each is growing every day. We are excited about the future and look forward to being a part of it. If you are looking for a quality group of DBAs to support your ever-changing data systems, check out our website and feel free to contact us. We would love to be a part of your future solutions.
Your Database Still Needs a Strategy
Automation can reduce routine work, but it does not replace the business judgment required to protect performance, control costs, secure data, and plan for what comes next. Moser Consulting’s database experts can help assess and support your evolving data environment.
