S13 Bonus: The AI DBA Shift: Automating Database Reliability & Performance with Itamar Syn-Hershko, Founder & CEO of NeverBlink

Episode Summary

Executive Summary: The episode tells the creation story of Neverblink AI, founded by Itamar Sinershko from a consulting practice that repeatedly rescued database customers in crisis. What began as scrappy scripts and a Grafana dashboard evolved into an AI-driven platform for monitoring, diagnosing, and improving database performance, first for OpenSearch/Elasticsearch and now expanding to ClickHouse. The conversation emphasizes iterative product-building, hiring, bootstrapping, and using real customer pain to shape the roadmap.

Main Topics: From consulting fires to product creation (Priority: 5/5): Neverblink AI originated as internal tooling built to solve urgent customer database incidents in the consulting business, then evolved into a standalone product as the same problems kept recurring. The scrappy MVP and iterative development (Priority: 5/5): The first version was a collection of scripts and a Grafana dashboard that monitored metrics the team couldn’t get from existing tools, then grew through repeated refinements into a platform. Trade-offs in early-stage engineering (Priority: 4/5): The team prioritized speed and customer relief over code quality and scale, accepting technical debt in order to ship quickly for high-pressure incidents. Bootstrapping, team quality, and hiring (Priority: 4/5): As a fully bootstrapped company, Neverblink focuses on hiring highly technical people who can solve hard problems and adapt across consulting and product work; the team is seen as the company’s strongest asset. Scaling databases and scaling the team (Priority: 4/5): The product itself is considered a solved scaling problem, but the company’s biggest challenge is scaling headcount and support capacity without VC-style overhiring. AI-era database reliability and expansion (Priority: 5/5): Neverblink is expanding beyond OpenSearch/Elasticsearch into ClickHouse and other databases, motivated by the rise of AI agents and non-deterministic workloads that stress data systems. Leadership lessons and advice (Priority: 3/5): Itamar stresses leading with business value, thinking far ahead, and helping others as a long-term operating principle for founders and builders.

Key Arguments: The product came directly from customer demand: urgent database outages forced the team to build tools fast, then reuse them as a platform. A scrappy MVP can be enough if it solves a real pain point; product maturity can come later through repeated iteration. Neverblink’s differentiation is 360-degree database visibility and root-cause correlation, not just surface-level infrastructure metrics. For a bootstrapped company, hiring only very strong people matters more than rapid headcount growth. Scaling the platform is easier than scaling the organization; human capacity and support load are the real bottlenecks. AI and agentic workflows increase database load and unpredictability, making database observability and performance management more important. Expanding from OpenSearch/Elasticsearch to ClickHouse is a natural extension of the company’s database reliability mission. Good founders should lead with business outcomes and think backward from long-term goals to near-term execution.

Data Points: Revenue milestone: $1 million ARR - Neverblink reached this after about 1.5 years of selling the tool as a product. Time to revenue milestone: 18 months - The product hit $1M ARR within 18 months of launch. Company funding model: Fully bootstrap - The company grew without VC funding. Founder workload: Father of five kids, all under seven - Used to illustrate Itamar’s personal time constraints outside work. Initial product scope: OpenSearch and Elasticsearch only - Neverblink initially focused on these search/database technologies before expanding. Newly launched support: ClickHouse support - The company recently expanded its platform to another database technology. MVP form: One big Grafana dashboard plus scripts - The first version of the product was described as scrappy and manually assembled.

Pivotal Quotes: "We basically built something that our market was demanding, right? Or not even market, the customers were demanding." — Itamar Sinershko: Explaining how the product emerged from repeated consulting emergencies. "Maybe sometimes building software is just evolving an MVP. Version 20, version 200, right?" — Itamar Sinershko: Reflecting on how products mature through continuous iteration rather than clean rewrites. "Always lead with the business value." — Itamar Sinershko: Advice to young entrepreneurs about how to present and build products.

Implications: The episode suggests that strong products often emerge from real operational pain, not abstract ideation. For database/AI infrastructure teams, it underscores the growing need for automated reliability tools and pragmatic, customer-driven iteration.