February 2026 Summaries

Many teams seek to unify their observability backends to avoid the complexity and costs of managing separate systems for logs, metrics, and traces. OpenTelemetry and Bindplane facilitate a smooth transition to ClickStack, allowing teams to integrate it alongside their current setups without disruption. This process involves adding ClickStack as an additional destination in existing telemetry pipelines, using routing logic to manage data flow, and employing Bindplane's processor capabilities to format and enrich data before it reaches ClickStack. ClickStack consolidates observability data into a single backend using ClickHouse, offering significant storage cost reductions due to its efficient columnar architecture. Migration strategies include running ClickStack in parallel with existing platforms, shifting specific workloads gradually, and employing filters and routing connectors to manage data flow. By turning unstructured logs into structured data, ClickStack enhances efficiency and speed, supporting a seamless transition that can typically be completed within weeks. Bindplane's support for multiple sources and destinations ensures adaptability to various systems, aligning with the industry's shift towards open standards and unified observability solutions.

The Agentic Data Stack is an open-source, composable architecture designed to integrate AI agents directly with your data, streamlining the analytics process by eliminating traditional steps like dashboard creation and ticket queues. It emphasizes data sovereignty, allowing users to maintain control over their data, conversations, and model choices. The stack comprises three main components: LibreChat, a versatile chat interface that supports multiple LLM providers; ClickHouse MCP, a middleware that enables rapid querying of large datasets; and Langfuse, which offers comprehensive observability and quality evaluation for LLM interactions. Organizations, including Shopify, Canva, and Khan Academy, are leveraging this architecture to enhance their data interactions, allowing non-technical users to query data effectively and explore insights autonomously. The stack is easily deployable on local machines using Docker, providing users with a quick setup experience and customizable options for user credentials.

ClickHouse offers a feature that allows users to create instant, zero-copy clones of production tables, enabling safe experimentation without affecting the original data. This process, akin to a Git fork, is made possible by ClickHouse's immutable data storage model, where every data insertion results in a new, self-contained part on disk. The cloning mechanism relies on hard-linking the column files from the source table to the cloned table, ensuring that the clone starts as an exact replica and diverges only upon modification. This part-level copy-on-write approach ensures that changes are isolated to specific parts, reducing storage overhead and maintaining high performance. The feature is particularly useful for testing destructive changes, schema experiments, and backfills, as it allows for rapid table replication independent of table size. In ClickHouse Cloud, cloned tables can be managed with isolated compute resources, further protecting the production environment while allowing users to explore the platform's capabilities with minimal risk.

ClickHouse, predominantly recognized as a high-performance analytics engine, has evolved to incorporate comprehensive geospatial functionalities over time. Initially, ClickHouse's geospatial capabilities were limited to basic coordinate-distance functions, but they have expanded significantly with the introduction of grid-based indexing systems like geohash and H3, and the integration of Google's S2 library. In 2025, a notable development was the introduction of a Geometry type, which allows for the storage of various geometry subtypes such as Point, LineString, and Polygon in a single column, enhancing the system's flexibility. ClickHouse now supports more than 90 geospatial functions, enabling operations like spatial indexing and polygon set operations, making it well-suited for large-scale spatial analytics tasks such as aggregating observations into grid systems and executing fast, indexed lookups. Despite its advancements, ClickHouse does not yet support traditional spatial indexing methods or coordinate reference system awareness, making it less suitable for projects requiring complex spatial joins or CRS-aware operations compared to specialized GIS tools like PostGIS. The recent enhancements reflect ClickHouse's ongoing commitment to improving geospatial analytics capabilities, making it a strong contender for managing large-scale location data efficiently.

LINEヤフー, a digital infrastructure giant born from the merger of LINE and Yahoo! Japan, utilizes ClickHouse and Apache Kafka to handle over a trillion messages and 2.6 petabytes of I/O daily, operating one of the world's largest Kafka deployments. The company has built a sophisticated observability stack for real-time analysis and debugging, which is essential for managing its extensive data transfer needs among various services, including messaging, news, and payment systems. ClickHouse, chosen for its SQL compatibility, compression capabilities, and performance, allows the team to efficiently query vast amounts of data, aiding in the detection and resolution of Kafka-related issues, such as race conditions and performance declines. This observability is crucial for maintaining system reliability at such a massive scale, enabling the company to quickly identify and address previously unknown problems, thereby contributing solutions back to the Kafka community.

The Postgres managed by ClickHouse service offers a robust, NVMe-backed Postgres database that is seamlessly integrated with ClickHouse, focusing on high availability, reliability, and operational observability for OLTP workloads. This enterprise-grade service features cross-availability zone high availability with synchronous standbys, ensuring durability and failover capabilities, along with a native integration for Change Data Capture (CDC) to replicate data into ClickHouse for real-time analytics. The service includes automatic backups with point-in-time recovery and employs WAL-G for efficient data archiving and retrieval. Security is prioritized through dedicated VPCs for network isolation, ensuring encrypted data both in transit and at rest, along with certificate-based client verification. Furthermore, the platform provides comprehensive monitoring and metrics, with plans to enhance performance insights and expand support for additional cloud providers. As part of its roadmap, the service aims to improve platform maturity with features like storage autoscaling, Terraform support, and seamless Postgres-to-Postgres migrations, reinforcing its position as a unified data stack for scalable applications.

Luzmo, a company specializing in embedded analytics for data-driven applications, transitioned from using open-source ClickHouse to ClickHouse Cloud to streamline infrastructure management and focus on building analytics solutions. This move enabled Luzmo to deliver high-performance, flexible, and cost-efficient analytics by integrating powerful visualizations and AI-driven features directly into applications, catering to diverse industries from marketing platforms to large-scale telecom operations. The company leverages ClickHouse for a variety of use cases beyond customer-facing analytics, including observability workloads and vector search, and has developed its own query engine to handle the platform's multi-tenant architecture and highly variable data schemas. ClickHouse Cloud's managed service allows Luzmo to offload operational tasks while retaining performance and scalability, helping them address challenges associated with unpredictable data environments and enabling fast, intuitive analytics experiences for end users.

DWAINE, an internal AI agent at ClickHouse, exemplifies how AI is reshaping business analytics by providing direct answers to stakeholders' queries, bypassing traditional dashboards. However, such AI systems can sometimes make errors due to inadequate context or suboptimal warehouse designs. The evolving landscape of data warehousing emphasizes the importance of isolating workloads by compute resources and exposing only curated data marts to AI to ensure accuracy and prevent misinformation. ClickHouse Cloud's architecture supports this by separating storage from compute, allowing dedicated resources for specific workloads without data duplication. The setup includes creating multiple services, such as ingestion and transformation services, and establishing access controls to implement guardrails. The transformation process involves organizing data into raw, staging, and marts layers, with a focus on canonical metrics and low latency. The article underscores the need for warehouses to meet the demands of conversational analytics while maintaining core principles to ensure reliable and accurate AI-powered insights.

Fivetran has launched a beta version of its destination for ClickHouse Cloud, introducing a significant new feature called History Mode, which supports Slowly Changing Dimension Type 2 by preserving every version of your data for comprehensive audit trails and point-in-time analysis. Collaboratively developed with Fivetran's engineering team, the connector allows seamless data integration from platforms like Salesforce, Segment, and Stripe into ClickHouse without the need for custom code. History Mode adds three system columns to track data versions, enabling powerful analytics such as change tracking and historical reporting. Additionally, the setup includes features like compute-compute separation, which isolates ingestion workloads, and optimization strategies such as skip indexes, projections, and materialized views to enhance query performance. The connector is designed to efficiently manage inserts, updates, and deletes in ClickHouse by automatically creating tables with primary keys or generating unique identifiers when necessary. Future updates aim to support advanced schema changes directly from the Fivetran dashboard to enhance usability and flexibility for users.

InMobi successfully transitioned its ad reporting system to ClickHouse to improve performance, cost-efficiency, and reliability, addressing the challenges of real-time decision-making for publishers. The migration from a serverless warehouse to ClickHouse reduced query latency from 60 seconds to under 3 seconds and cut monthly expenses by 80%, all while handling over 400,000 daily queries across 10 terabytes of data. The new architecture uses Spark for data ingestion and a parent-child service pattern for efficient resource management, effectively addressing issues such as atomic ingestion, concurrency, and fallback mechanisms. This transformation has not only enhanced publisher reporting but also established a scalable, predictable analytics platform for future use cases. By implementing these changes, InMobi improved its service reliability and gained the ability to forecast costs more accurately, which is crucial given the company's anticipated growth.

ClickHouse 25.12 introduces an HMAC function for verifying webhook requests, enhancing its utility as a webhook endpoint by filtering out requests from unauthorized sources. The process involves setting up a series of tables and materialized views to authenticate, transform, and store verified data, utilizing ClickHouse's SQL capabilities. A staging table initially receives data, which is validated and transformed by a materialized view before being stored in a final table. To demonstrate, a local example is created using a restricted user for security, and the setup is then expanded to a real-world application with GitHub webhooks in ClickHouse Cloud. This involves configuring tables to handle events and signatures, ensuring only data with valid HMAC signatures is logged. The guide also highlights the importance of secure user permission settings and the flexibility of ClickHouse's HTTP insert capabilities.

In February 2026, ClickHouse made significant advancements, including raising $400 million in a Series D funding round and acquiring Langfuse, an open-source platform for large language model observability. They also introduced an enterprise-grade PostgreSQL service and an official Kubernetes Operator to simplify ClickHouse cluster deployment. The newsletter highlights Ino de Bruijn's contributions to cancer research through data visualization and a conversational AI interface for cBioPortal using ClickHouse. Upcoming global events and workshops are announced, focusing on real-time analytics, query optimization, and data warehousing. The new 26.1 release includes various enhancements, such as the sparseGrams tokenizer and support for the Variant data type. Additionally, the newsletter discusses innovative solutions for ensuring data completeness in cryptocurrency trading and optimizing data models for real-time analytics with ClickHouse. The combination of PostgreSQL and ClickHouse is promoted as a unified open-source data stack, offering seamless integration for transactional and analytical workloads.

PeerDB offers a robust solution for large-scale Postgres to Postgres migrations, emphasizing minimal downtime and high-speed data transfer, as demonstrated in its comparison with pg_dump/pg_restore and native logical replication. This tool is particularly effective for online migrations, offering high initial load throughput and continuous change data capture (CDC) while supporting complex PostgreSQL workloads, including large TOAST columns and schema evolution. PeerDB's architecture enables parallel processing for data transfers, significantly reducing initial load times compared to traditional methods. By leveraging PostgreSQL’s binary format and logical partitioning, PeerDB ensures efficient data migration with preserved data fidelity, even for complex data types, while providing enhanced observability and automatic failure retries to mitigate risks during migrations. Its open-source nature allows for flexibility across different setups, including self-hosted and cloud environments, positioning PeerDB as a versatile tool for teams requiring seamless and reliable database migration solutions.

In December 2025, the PostgreSQL extension pg_clickhouse was launched to facilitate seamless querying of ClickHouse directly from Postgres, aimed at reducing the migration effort for analytics workloads from Postgres to ClickHouse. By offloading analytic execution to ClickHouse, pg_clickhouse minimizes the load on Postgres, allowing for independent scaling and avoiding resource contention, particularly beneficial for aggregation-heavy queries. The extension is evaluated using ClickBench, where it demonstrated superior performance among PostgreSQL extensions, closely matching the speed of native ClickHouse. This performance is achieved through effective query pushdown, which handles a wide range of operations, and the extension continues to evolve with plans to support more complex queries. Pg_clickhouse is available as an open-source option and within the managed Postgres service in ClickHouse Cloud, marking a strategic choice for scalable applications.

Migrating analytical workloads from Postgres to ClickHouse can be significantly accelerated using AI, provided the right environment is established for AI agents to succeed. The integration of Postgres and ClickHouse offers a unified data stack that leverages the transactional strengths of Postgres and the analytical capabilities of ClickHouse. To effectively utilize AI in this process, organizations should treat their data stack as code, enabling AI agents to operate within familiar coding environments and allowing for easy rollbacks and iterations. The MooseStack framework is highlighted as a tool to facilitate this migration, offering fast feedback loops and integration with best practices for ClickHouse, enabling agents to iteratively refine their approach and align with OLAP-native patterns. By providing AI agents with comprehensive context, including schemas, skills, and reference implementations, they can effectively translate existing business logic into optimized OLAP models, transforming the migration into a manageable, code-driven process that is both scalable and reliable.

SmartNews has transformed its ad performance analytics by migrating from a batch-processing system using Amazon Redshift to a real-time streaming architecture powered by ClickHouse, significantly improving query response times for advertisers on Japan's leading news app. This shift allows SmartNews to handle approximately 3 million queries and 60,000 data inserts per hour, with most queries completing in under 100 milliseconds. The transition involved replacing a multi-step batch process with a streamlined pipeline where ad delivery data is streamed through Kafka and processed by Flink before being ingested into ClickHouse. This new system allows advertisers to access real-time metrics such as budget spend and click-through rates almost instantaneously. SmartNews has also scaled its infrastructure by migrating from AWS EC2 to Kubernetes, enhancing flexibility and scalability. The use of ClickHouse’s uniqCombined function allows SmartNews to efficiently estimate ad reach by counting unique users through a compressed intermediate state, maintaining accuracy even as the number of unique users scales. This investment in ClickHouse has not only improved performance and simplicity but also provided a flexible foundation for future growth and innovation in their advertising platform.

Cloudflare's approach to scalability, as highlighted by Jamie Herre, emphasizes the inevitability of system failures and the necessity of designing infrastructure that can adapt and remain resilient under stress. At a ClickHouse meetup, Jamie demonstrated the robustness of Cloudflare's analytics system, capable of processing quadrillions of events and maintaining performance even under severe conditions, such as simulated data center outages. This resilience is attributed to Cloudflare's use of ClickHouse, an open-source OLAP database that offers simple integration, minimal coordination, and the flexibility to handle extreme scales without complex tradeoffs. Jamie underscores the importance of preparing for scalability challenges proactively, suggesting that the principles of designing for both explosive growth and potential failures are applicable regardless of company size, with ClickHouse serving as a crucial tool in their strategy.

pg_stat_ch is a newly open-sourced PostgreSQL extension that records each query execution as a 4.6KB event and streams it into ClickHouse for detailed analysis, akin to application performance monitoring. It supports PostgreSQL versions 16 to 18 and is designed to provide deep insights into query behavior, including latency, top queries, and error tracking, which are not available out-of-the-box with PostgreSQL. The extension captures comprehensive query data, minimizes overhead on PostgreSQL, and prioritizes nonintrusive operation by avoiding back-pressure, ensuring that PostgreSQL performance remains unaffected even if ClickHouse slows down. The architecture uses fixed-size events to maintain predictable memory usage and employs efficient data handling methods, such as LZ4 compression, to minimize storage and network bandwidth. While pg_stat_ch is not yet in production, it seeks feedback from users operating at scale to refine its capabilities. It offers a complementary solution to existing tools, like pg_stat_statements and pg_stat_monitor, by providing unlimited retention and raw event granularity, allowing for sophisticated queries and analyses using ClickHouse's SQL capabilities. The extension is part of a broader strategy to integrate PostgreSQL with ClickHouse, forming a unified data stack for scalable applications, and future enhancements include query plan capture, sampling support, and pre-built dashboard templates.

Metrics have traditionally played a crucial role in monitoring systems by providing quick insights into system health through pre-aggregated statistics that are cheap to store and query, particularly in environments with high data volumes. However, their limitations in handling high-cardinality data and lacking correlation with logs and traces have become apparent, often necessitating deeper dives into more detailed data sources for effective troubleshooting. As a result, there's a growing shift towards using structured events and traces, especially with advancements in columnar databases like ClickHouse, which allow for efficient storage and analysis of high-fidelity data, transforming metrics from a primary monitoring tool to a performance optimization layer. This evolution in observability tools enables more flexible investigations, allowing teams to define metrics post-incident rather than preemptively, aligning with the broader trend of moving from ETL to ELT processes, and paving the way for more dynamic, data-driven decision-making in system monitoring.

In January, ClickStack, an open-source observability stack for ClickHouse, introduced significant updates, including the launch of Managed ClickStack, which offers a fully managed observability solution on ClickHouse Cloud, optimizing costs by charging for compute and storage rather than events or users. The improvements in index support, particularly with Bloom filters and the introduction of inverted indices, have significantly enhanced query performance for large datasets. ClickStack has also integrated materialized views to optimize alerts, making them more efficient by shifting aggregation work to the ingestion phase. Additional updates include support for the SQL HAVING clause in queries, legend filtering for charts, and improvements to the OpenTelemetry Collector for easier onboarding and independent operation. A new ClickStack theme, aligning with ClickHouse branding, was introduced alongside these technical advancements, reflecting ongoing efforts to enhance usability and scalability.

ClickHouse version 26.1 introduces a host of new features, performance optimizations, and bug fixes, including significant improvements in asynchronous insert deduplication with materialized views, new syntax for indexing projections, and support for the Variant type across all functions. The release enhances the capabilities of ClickHouse's analytics framework by addressing previous limitations in deduplication with materialized views, ensuring retry safety in complex ingestion pipelines, and optimizing operations on LowCardinality columns. Additionally, introspection and diagnostics for MergeTree tables and the Keeper state are improved with new system tables and functions, while the release also brings advancements in text indexing and the inclusion of a web UI and HTTP interface for Keeper. The ClickHouse community welcomes new contributors, and the release emphasizes the importance of community contributions in evolving the platform. Furthermore, a new Kubernetes operator is introduced, providing open-source support for automated cluster provisioning and management, enhancing usability and scalability for users deploying ClickHouse in containerized environments.

Wix, a global platform for digital presence, manages a highly complex production environment with over 300 million users and 4,000+ microservices, necessitating robust incident response mechanisms. To enhance the speed and accuracy of incident response, Wix turned to AI-driven solutions, leveraging its existing infrastructure and zero-error culture. The company integrated ClickHouse, a high-performance columnar database, for efficient log management and real-time analytics, which proved essential for their AI-driven system, Wild Moose. Wild Moose, an AI platform designed to mimic the investigative processes of engineers, automates incident response by processing vast amounts of data quickly, providing enriched alerts and reducing manual workload. This system allows engineers to start with context and probable root causes, improving accuracy and reducing mean time to resolution (MTTR), which in turn boosts team morale and decreases on-call stress. By combining AI, observability, and a strong data infrastructure, Wix has created a scalable, AI-driven reliability system that transforms tribal knowledge into institutional knowledge, thereby maintaining their zero-error tolerance while continuing to scale their platform.

LY Corporation utilizes ClickHouse to manage and observe its vast Kafka deployment, which processes over 1 trillion messages and 2.6 petabytes of data daily. With 24 servers, LY processes 7 million rows per second of API logs for real-time debugging, helping engineers address unique challenges that arise from operating at such a large scale. Born from the merger of LINE and Yahoo! Japan, LY operates a wide array of digital services, including Japan's most popular messaging app and a significant news portal, all interconnected by a massive Kafka platform. This system handles 31 million messages per second, requiring advanced observability to maintain performance and resolve unprecedented issues. LY's observability stack, resembling a research lab, leverages ClickHouse for its SQL compatibility, compression capabilities, and performance, enabling efficient data management and analysis. Engineers use ClickHouse's queryability to diagnose and solve complex Kafka bugs, like a race condition affecting message offsets. ClickHouse's integration into LY’s observability platform provides crucial real-time forensics, allowing the team to maintain one of the world’s largest Kafka systems effectively.

A comprehensive benchmark analysis comparing five major cloud data warehouses—Snowflake, Databricks, ClickHouse Cloud, BigQuery, and Redshift—was conducted at varying scales of 1 billion, 10 billion, and 100 billion rows to evaluate their cost-performance metrics. The study revealed that ClickHouse Cloud consistently outperformed all other systems in terms of cost efficiency across all scales, maintaining a "fast & low-cost" status even as data volumes increased significantly, while other systems shifted towards "slow & high-cost" zones. The results were generated using Bench2Cost, an open-source benchmark pipeline that applies actual compute billing models to raw execution times for accurate cost comparisons, excluding storage costs due to their negligible impact. The study incorporated an interactive benchmark explorer to analyze various configurations, with the findings showing ClickHouse Cloud as the clear leader, offering exponentially superior cost performance compared to its competitors, which struggled with higher costs and slower execution times as data scaled up. The benchmark methodology employed standard workloads without engine-specific optimizations, ensuring fair and reproducible results, and highlighted ClickHouse Cloud's scalability without encountering hard limitations faced by other systems.

Silverflow has revolutionized its data analytics by implementing ClickHouse Cloud, significantly reducing latency and streamlining its data infrastructure to support near-real-time insights on global payment events. This transition has allowed Silverflow to cut end-to-end latency for data insights from 20 minutes to just 65 seconds and replace numerous ETL pipelines with ClickHouse’s dynamic JSON ingestion and ClickPipes, which has reduced engineering costs and complexity. Founded to modernize card-processing infrastructure with a unified API for direct network connections, Silverflow has expanded globally since its 2019 launch, necessitating robust data solutions to handle increased volume and complexity. By integrating ClickHouse Cloud, Silverflow has simplified its data ingestion into a single table with materialized views, enhancing performance and enabling timely data access for real-time dashboards and decision-making. As the company continues to grow, it plans to further integrate ClickHouse, decommission legacy ETL pipelines, and explore new capabilities like regional deployments to meet data residency requirements, positioning itself for continued agility and competitive advantage in the fast-paced payments industry.

Managed ClickStack for ClickHouse Cloud has been announced in beta, offering a streamlined approach to utilizing ClickHouse for observability by providing fast queries and powerful analytics across various data types without the need for managing infrastructure. Emphasizing cost efficiency, it allows users to store high-cardinality OpenTelemetry data at a significantly reduced price, scaling logs, metrics, and traces independently and eliminating traditional constraints like reduced retention periods and sampling. The platform fosters an economic shift in observability, enabling long-term data retention without hefty costs, and is built on a foundation that supports separation of storage and compute, allowing users to pay only for what they use. Managed ClickStack aims to unify observability data with other business and application datasets in a single platform, facilitating deeper insights and correlations, thus setting the stage for future advancements in agentic workloads and AI-driven data interaction. Existing ClickStack Open Source users can seamlessly transition to this managed service, and new users can begin with a ClickHouse Cloud trial, which includes ample credits to explore the benefits of the platform. As the beta progresses, the offering will expand and refine, with ongoing community engagement encouraged to shape its development.

The official ClickHouse Agent Skills have been released as a set of open-source tools encoded with best practices for schema design, query optimization, and data ingestion, aiming to enhance the functionality of AI assistants through 28 prioritized rules. These skills are designed to transform general-purpose language models into ClickHouse power users by addressing common issues such as inefficient ORDER BY choices, non-scalable JOINs, and improper use of Materialized Views or specialized indexes. The skills are packaged in an agent-agnostic format, making them easily adoptable for developers who can add them to their local environments or invoke them manually. The initiative, which encourages community contributions under an Apache license, seeks to support developers in overcoming challenges posed by less-than-perfect AI-generated solutions and plans to expand into more complex areas like cluster configurations and engine-specific optimizations.

Organizations often face difficulties when trying to analyze relationships in large datasets using SQL due to limitations in handling complex queries, leading to considerations of using graph databases. However, this typically requires copying data into a graph database, creating issues with data sync, dual storage costs, and maintenance burdens, which can be both costly and inefficient. The zero-copy graph analytics approach provides a solution by allowing graph queries directly on existing data, integrating a graph query layer like PuppyGraph on top of an analytical database such as ClickHouse. This method removes the need for data movement, offering real-time data analysis, reducing storage costs, and simplifying system maintenance. By implementing this approach, companies can efficiently perform multi-hop relationship queries and analytical tasks without sacrificing performance, as demonstrated by use cases in customer recommendations and fraud detection. This shift towards integrating graph queries within existing data architectures reflects a broader trend in data management, emphasizing unified systems that can handle diverse analytical needs without duplicating data.

Langfuse leverages ClickHouse Cloud to enhance real-time observability and analytics for large-scale, complex language model (LLM) applications, transitioning from Postgres to ClickHouse for improved performance and scalability. This transition significantly reduced query latency and supported increased data ingestion volumes, leading to a shift towards immutable, wide-table modeling that optimizes memory usage and query speed. As LLM applications evolved from simple retrieval-augmented generation (RAG) workflows to complex agentic systems, Langfuse adapted its data model, ingestion patterns, and SDK design to meet the demands of these advanced systems. Observability challenges shifted focus from basic request success to understanding agent decision-making processes, prompting Langfuse to expand from trace trees to graph views and analytics tools, allowing teams to analyze latency, cost, and quality of millions of runs. Langfuse's move to ClickHouse, facilitated by an extensive pipeline redesign, addressed previous scalability issues and introduced a more adaptable, open-source data platform. The migration allowed for performance improvements while maintaining flexibility in user data ingestion and analysis. The company is now transitioning to an immutable data model, further enhancing performance and aligning with changing user expectations, while also backfilling legacy data to accommodate existing deployments. The recent acquisition of Langfuse by ClickHouse further integrates their capabilities, continuing the journey of scaling observability for agentic systems.

ClickHouse Inc., a leading real-time analytics data platform company, has appointed Takeshi Kaneko as the President of its Japanese subsidiary, ClickHouse K.K., effective February 1, 2026. Kaneko brings over 20 years of experience in enterprise business leadership from global companies such as Microsoft Japan and Red Hat, and he previously served as the Corporate Vice President and President at Nutanix Japan. ClickHouse, known for its high-speed, column-oriented database system designed for real-time data processing and analysis, has rapidly expanded globally since its 2016 launch, with over 3,000 companies worldwide using its platform, including major firms like Microsoft, Sony, and Netflix. The company emphasizes its platform's ability to handle complex and frequent queries with high throughput and low latency, making it ideal for AI-driven decision-making. Kaneko aims to leverage his experience to strengthen ClickHouse's presence in Japan, assisting local businesses in transforming AI and data into competitive advantages.