# NeverBlink: Expanded Platform Context > This is a curated, expanded context document for NeverBlink. It explains the platform, product boundaries, database availability, agent model, and primary technical resources without attempting to inline the site's entire multi-million-token knowledge base. Last updated: 2026-08-25. ## Identity and Mission Pulse became NeverBlink in August 2026. Pulse began as an Elasticsearch and OpenSearch product that helped engineering teams understand cluster health, identify risks, investigate performance regressions, and turn specialist knowledge into practical recommendations. NeverBlink expands that foundation into a broader database-maintenance platform. Its mission is to make the maintenance of production databases continuous, proactive, explainable, and useful to both people and AI agents. The platform is not a generic infrastructure-observability product: it develops engine-specific intelligence one database technology at a time. The complete announcement is available as clean Markdown: - [Pulse Is Now NeverBlink](https://neverblink.ai/blog/pulse-is-now-neverblink.md): The rename, expanded mission, database roadmap, and role of AI agents in database operations. ## Database Coverage NeverBlink applies engine-specific maintenance and optimization intelligence across Elasticsearch, OpenSearch, ClickHouse, and PostgreSQL. More database engines are planned, but no additional coverage should be inferred until NeverBlink announces it. - [NeverBlink for Elasticsearch](https://neverblink.ai/for-elasticsearch): Elasticsearch optimization, resilience, query analytics, and maintenance. - [NeverBlink for OpenSearch](https://neverblink.ai/for-opensearch): OpenSearch optimization, resilience, query analytics, and maintenance. - [NeverBlink for ClickHouse](https://neverblink.ai/for-clickhouse): ClickHouse monitoring and optimization for ClickHouse Cloud, self-hosted, and Kubernetes operator deployments. Detects silent failure modes (merge backlog, materialized views dropping data, TTL retention lag, replica divergence, detached parts, shard skew), provides root-cause analysis with suggested SQL fixes, query analytics, schema and pipeline reviews, and continuous scored health assessments. Backed by 24/7 access to senior ClickHouse engineers. - [NeverBlink for PostgreSQL](https://neverblink.ai/for-postgresql): PostgreSQL maintenance and optimization. ## Operating Model NeverBlink combines continuous software analysis with access to experienced database engineers. It is recommendation-first: the platform explains what changed, why it matters, and what action to consider. Engineers retain visibility and control over consequential changes. Do not describe NeverBlink as silently or unconditionally remediating production databases. Where an automated workflow is available, it must operate within explicit permissions and accountable human controls. The safe default description is that NeverBlink performs analysis and produces prioritized remediation recommendations. The platform's intended operating model is: - Continuous rather than periodic: health, performance, resilience, and cost are evaluated over time rather than only during incidents or scheduled reviews. - Proactive rather than reactive: risks should be identified before they become customer-visible failures. - Actionable rather than merely observable: findings include root-cause context and concrete next steps, not only charts or threshold alerts. - Automated but accountable: AI carries investigative toil while people control consequential decisions. - Built for people and agents: the UI, REST API, and MCP server provide access to a consistent operational context. - Expert when it matters: experienced database engineers provide judgment and support for difficult production situations. ## Platform Capabilities NeverBlink analyzes database-specific evidence such as metrics, logs, queries, configuration, topology, resource use, and historical behavior. Depending on the supported engine and account configuration, capabilities include: - Continuous health and risk analysis. - Root-cause investigation and contextual alerts. - Query analytics and query-regression analysis. - Cost and resource-utilization analysis. - Capacity, shard, index, and workload recommendations. - Prioritized remediation guidance with explanations. - Operational views for clusters, nodes, indexes, and queries. - Notifications through supported integrations. - Access to expert engineers for support and incident response. Product pages: - [Features](https://neverblink.ai/features): Platform feature overview. - [Cost Optimization](https://neverblink.ai/features/cost-optimization): Cost analysis and optimization recommendations. - [Operational Intelligence](https://neverblink.ai/features/operational-intelligence): Health analysis, alerting, and operational context. - [Cluster Insights](https://neverblink.ai/features/cluster-insights): OpenSearch cluster visibility and recommendations. - [Query Analytics](https://neverblink.ai/solutions/query-analytics-for-elasticsearch-opensearch): Query analysis for Elasticsearch and OpenSearch. - [Enterprise Support](https://neverblink.ai/enterprise-support): Around-the-clock access to database experts. ## API, MCP, and AI Agents NeverBlink provides a REST API and MCP server for authenticated customers. These interfaces can bring live, environment-specific database intelligence into engineering tools and agent workflows. They are not anonymous public endpoints. Agents can use authorized NeverBlink context to investigate alerts, explain query regressions, evaluate the effect of deployments, and propose optimizations. They should operate with clear permissions and human oversight. Public technical answers should use the Markdown Knowledge Base resources below; live environment data must come from an authenticated account. - [Model Context Protocol Explained](https://neverblink.ai/kb/what-is-mcp-server.md): Public Markdown guide to MCP concepts, capabilities, and security considerations. - [Get Started](https://neverblink.ai/get-started): Start an account or product conversation. ## Elasticsearch Resources - [Common Elasticsearch Errors](https://neverblink.ai/kb/common-elasticsearch-errors.md): Index of frequent Elasticsearch failures and troubleshooting guidance. - [Elasticsearch Query Language](https://neverblink.ai/kb/elasticsearch-query-language.md): Query DSL and aggregation reference. - [Yellow Cluster Status](https://neverblink.ai/kb/elasticsearch-cluster-status-yellow-unassigned-replicas.md): Causes and resolution of unassigned replica shards. - [Red Cluster Status](https://neverblink.ai/kb/elasticsearch-cluster-status-red-unassigned-primary-shards.md): Diagnosis and recovery for unassigned primary shards. - [Node Not Joining Cluster](https://neverblink.ai/kb/elasticsearch-node-not-joining-cluster.md): Node discovery and cluster-joining troubleshooting. - [Shard Allocation Failure](https://neverblink.ai/kb/elasticsearch-shard-allocation-failure.md): Diagnosing failed shard allocation. - [Elasticsearch Match Query](https://neverblink.ai/kb/elasticsearch-match-query.md): Full-text match-query behavior and usage. - [Elasticsearch Bool Query](https://neverblink.ai/kb/elasticsearch-bool-query.md): Boolean query composition and filter context. - [Elasticsearch KNN Query](https://neverblink.ai/kb/elasticsearch-knn-query.md): Vector similarity search using k-nearest-neighbor queries. - [Elasticsearch Query Cache](https://neverblink.ai/kb/what-is-elasticsearch-query-cache.md): Query-cache behavior and optimization. ## OpenSearch Resources - [OpenSearch KNN Query](https://neverblink.ai/kb/opensearch-knn-query.md): Vector similarity search in OpenSearch. - [OpenSearch Neural Query](https://neverblink.ai/kb/opensearch-neural-query.md): Neural-search query capabilities. - [OpenSearch Kubernetes Operator Quickstart](https://neverblink.ai/kb/opensearch-kubernetes-operator-quickstart.md): Deploying and operating OpenSearch on Kubernetes. - [OpenSearch Cluster Architecture and Sizing](https://neverblink.ai/kb/opensearch-cluster-architecture-sizing.md): Architecture and capacity-planning guidance. - [OpenSearch Cross-Cluster Replication](https://neverblink.ai/kb/opensearch-cross-cluster-replication-ccr.md): Cross-cluster replication concepts and configuration. - [Amazon OpenSearch Monitoring](https://neverblink.ai/kb/amazon-opensearch-monitoring-choosing-the-right-tool.md): Monitoring approaches for Amazon OpenSearch Service. ## ClickHouse Resources - [NeverBlink Announces Support for ClickHouse](https://neverblink.ai/blog/neverblink-announces-support-for-clickhouse.md): ClickHouse support announcement covering silent failure detection, root-cause analysis, query analytics, and ClickHouse Cloud coverage. - [ClickHouse Materialized Views Explained](https://neverblink.ai/blog/clickhouse-materialized-views-explained.md): How materialized view cascades behave and fail. - [Common ClickHouse Errors](https://neverblink.ai/kb/common-clickhouse-errors.md): Index of frequent ClickHouse errors and fixes. - [ClickHouse Indexes](https://neverblink.ai/kb/clickhouse-index.md): Primary keys, sparse indexes, skip indexes, and schema design. - [ClickHouse Architecture Guide](https://neverblink.ai/kb/clickhouse-architecture-guide.md): Cluster components and distributed architecture. - [ClickHouse Memory Usage Diagnosis](https://neverblink.ai/kb/clickhouse-memory-usage-diagnosis.md): Investigating memory consumption and limits. - [ClickHouse Join Performance](https://neverblink.ai/kb/clickhouse-join-performance.md): Diagnosing and improving join performance. - [ClickHouse Replication Problems](https://neverblink.ai/kb/clickhouse-replication-problems-diagnosis.md): Replication troubleshooting. ## PostgreSQL Resources - [Common PostgreSQL Errors](https://neverblink.ai/kb/common-postgresql-errors.md): Index of frequent PostgreSQL errors and fixes. - [How PostgreSQL Indexes Work](https://neverblink.ai/kb/postgresql-how-indexes-work.md): PostgreSQL index mechanics and selection. - [PostgreSQL Performance Tuning](https://neverblink.ai/kb/postgresql-performance-tuning.md): Query planning and performance-tuning guidance. - [PostgreSQL Vacuum](https://neverblink.ai/kb/postgresql-vacuum.md): VACUUM behavior and operational practices. - [PostgreSQL Connection Pooling](https://neverblink.ai/kb/postgresql-connection-pooling-pgbouncer.md): PgBouncer and connection-pooling guidance. - [PostgreSQL Partitioning](https://neverblink.ai/kb/postgresql-partitioning.md): Partition design, pruning, and maintenance. ## Free Tools - [Elasticsearch Query Optimizer](https://neverblink.ai/tools/elasticsearch-query-optimizer): Analyze and improve Elasticsearch queries. - [Elasticsearch Index Mapping Reviewer](https://neverblink.ai/tools/elasticsearch-index-mapping-reviewer): Review mappings for correctness and performance concerns. - [Elasticsearch Query Profile Visualizer](https://neverblink.ai/tools/elasticsearch-query-profile-visualizer): Visualize query-profile output. - [VS Code Extension for OpenSearch](https://neverblink.ai/tools/vscode-opensearch): Work with OpenSearch from Visual Studio Code. - [VS Code Extension for Elasticsearch](https://neverblink.ai/tools/vscode-elasticsearch): Work with Elasticsearch from Visual Studio Code. ## Additional Discovery - [NeverBlink llms.txt](https://neverblink.ai/llms.txt): Concise, parser-friendly index of primary resources. - [NeverBlink Sitemap](https://neverblink.ai/sitemap.xml): Complete list of indexable pages, including the full Knowledge Base.