Rootconf
Rootconf

Rootconf topical edition on Databases

Call for Proposals

All submissions

Shivek Khurana

@shivekkhurana

How to use SQLite in production and not get fired

Submitted May 25, 2026

Session type - select the format for your session: 30-minute talk – technical deep dive

Introduction

If your app uses a relational datastore, the default choice is usually Postgres or another client-server database. That is the accepted practice, but it also adds real system complexity.

As application developers, we mostly see SQL queries. From an infrastructure perspective, however, we are now running a large database server with its own operational surface area: networking, security, I/O tuning, backups, replication, monitoring, and failure modes.

SQLite simplifies this model. Instead of running a database server and an application server that communicate over TCP, the database is embedded directly inside the application.

The usual objection is: “SQLite is slow and does not scale.” But SQLite is more than two decades old, still actively maintained, and still backward compatible with its earliest APIs. If you use SQLite like it is 20 years old, it will not scale well. But modern SQLite, when configured properly, can handle thousands of concurrent readers and many production workloads, as long as you understand its write-concurrency model.

This talk debunks common SQLite myths and shows how to tune it for real production use.

Takeaways

  • Understand modern SQLite APIs and production patterns
  • Compare the complexity of networked databases with the simplicity of embedded SQLite
  • Tune SQLite for concurrency, caching, backups, and workload-specific performance
  • Use monitoring to understand and tune bespoke workloads
  • If time permits: use SQLite as a lightweight vector store

Audience

  • DevOps and infrastructure managers
  • Solo developers and vibe-coders
  • Indie hackers who want a simple stack and faster path to PMF
  • Engineers in larger teams who own small services or microservices
  • Builders who believe vertical scaling is still underrated

Bio

Shivek Khurana is an entrepreneur, angel investor, and software developer. He started building apps in 2008 with PHP and later became a Clojure developer.

As an engineer, he has built cybersecurity systems and machine learning pipelines for companies across the UK, Brazil, and the Nordic region. As a founder, he built Meta Blocks, a Solana upgrade protocol that reached approximately $10M in value and became profitable within two years.

After shutting down his company, he shifted focus toward AI agents. He currently works with a US-based VC firm, building internal AI tools and research pipelines.

Outside work, he invests in ed-tech and health-tech, writes on his blog, and experiments with fashion design.

References

All submissions

Comments

{{ gettext('Login to leave a comment') }}

{{ gettext('Post a comment…') }}
{{ gettext('New comment') }}
{{ formTitle }}

{{ errorMsg }}

{{ gettext('No comments posted yet') }}

Hosted by

Rootconf
Rootconf
We care about site reliability, cloud costs, security and data privacy