Loading…
Latest reads
Every post, summarised
Grounded summaries of engineering blogs, newest first. Each links to the original.
- 4,671 posts
- 20 companies
The Good Stuff
Paid picks we use ourselves. Links carry a referral code — a subscription through one earns us a commission at no extra cost to you.
Latest posts
Getting started with clickhousectl: the ClickHouse CLI for local and cloud
ClickHouse developers often face friction managing local versions, isolated instances, project scaffolding, and Cloud operations without relying on complex shell scripts. The open-source command-line tool clickhousectl addresses this by unifying local environment management and ClickHouse Cloud workflows. Locally, it downloads shared binaries, initializes standardized directory structures, and spins up named, isolated servers with automatic port assignments to facilitate upgrade testing. For cloud workflows, the CLI enforces a split authentication model where browser OAuth provides read-only access and API keys permit state mutations like service creation and scaling. Furthermore, the tool includes Claude Code agent skills covering best practices and deployment workflows, alongside support for CI/CD pipeline automation via environment variables and JSON output.
Introducing the Databricks Connector for Google Sheets: Real-Time, Governed Lakehouse Data in the Sheets Users Love
Databricks has introduced a Google Sheets connector that brings live, governed lakehouse data into a familiar spreadsheet environment. The connector addresses reliance on CSV exports, snapshots, and custom pipelines, which left business teams with siloed copies, inconsistent metrics, and slower decisions. Built on Databricks SQL and Unity Catalog, it lets users query governed datasets, including Unity Catalog Metric Views, through a no-code GUI or SQL, while Unity Catalog manages permissions. Users can refresh results manually or schedule refreshes, keeping Sheets data current without exporting snapshots. The connector is Generally Available to all Databricks customers, and Nubank reports that it has enabled non-technical users to explore data in Sheets while preserving control.
What is operational analytics?
Operational analytics uses real-time or near-real-time data to monitor day-to-day operations and support immediate decisions, unlike traditional analytics, which generally relies on batch data to explain past performance. Its workflow collects logs, events, clickstreams, telemetry, transactions, and other signals; streams them into a unified platform for cleaning, transformation, and enrichment; analyzes incoming data with rules, models, or anomaly detection; and sends results to dashboards, alerts, or operational applications. This can help teams detect issues earlier, reduce mean time to detect (MTTD) and mean time to respond (MTTR), improve forecasting, coordinate across departments, and act on inventory, customer, or system changes. The approach also requires integrating heterogeneous systems, maintaining data quality and schemas, embedding insights into existing workflows, and building reliable low-latency pipelines, with tools such as Lakeflow, Databricks SQL, and AI and machine learning capabilities presented as examples.
Frontend Engineering at Palantir: Polar Scaled Tiles in Zodiac
Zodiac, Palantir’s custom 3D globe library built on Three.js, encountered severe performance degradation near the Arctic because latitude/longitude tiles converge into tiny slivers and create excessive geometry. Standard equal-angular tiling was replaced internally with polar-scaled tiles whose widths expand toward the poles so their surface areas remain closer to equatorial tiles. The design uses spherical surface-area calculations, power-of-two merge spans, and dynamic quadtree expansion; client-side texture compositing handles standard geodetic or Web Mercator sources, while heightmaps are combined and resampled for wider tiles. The change reduced polar geometry, draw calls, and textures by over 90% and made frame time more consistent, though texture loading remains the next bottleneck and native polar tiles would remove stitching overhead.
AI Prototyping Is Changing How We Build Products at Uber
Uber describes how AI-assisted prototyping is changing product development by making interactive flows available within hours instead of weeks of coordination. In one example, a Merchant team customized a prototype to a specific catalog, producing immediate actionable feedback and rapid internal alignment. Across experiments, teams used Lovable, Figma Make, Claude Code, and Cursor to explore multiple concepts before PRDs and architecture decisions hardened; one PM explored six concepts in about 20 minutes. The pattern was earlier exploration, faster alignment, and clearer MVP boundaries, while PRDs and design reviews remained necessary for intent, tradeoffs, constraints, and durable decisions. Uber also opened prototyping to engineering and operational teams, while examining how prototypes should inform production without being mistaken for shippable solutions.
Unity AI Gateway: How to connect agents to external MCPs securely
Databricks Unity AI Gateway provides a governed way to connect agents to external Model Context Protocol (MCP) servers while integrating management with Unity Catalog. The post addresses provider-specific OAuth app registration, client secrets, token refresh, permission auditing, and the lack of centralized visibility into agent access. It describes registering an external MCP server, choosing per-user OAuth or a shared principal, selecting GitHub, and letting Databricks handle OAuth registration, token exchange, and refresh; connections can then be tested in AI Playground or with DatabricksMCPClient and deployed through Agent Bricks. Unity Catalog supplies permissions and audit records, while MLflow Tracing captures requests and tool interactions, and supported providers include Glean, GitHub, Atlassian, Google Drive, and SharePoint across AWS, Azure, and GCP.
Expanding agent governance with Unity AI Gateway
Unity AI Gateway extends Unity Catalog’s permissions, auditing, and policy controls to agentic workflows involving LLMs, MCP servers, and APIs. The release addresses limited visibility across multi-step actions with on-behalf-of user execution for MCP calls, configurable guardrails, and consistent governance across model providers without separate configurations for each provider and model choice can vary by task. Guardrails can detect PII, prompt injection, unsafe content, and data exfiltration, while logs capture identities, timestamps, MCP details, dollar costs, payloads, latency, status codes, and errors. OpenAI-compatible APIs, fallback models, and endpoint-, user-, or group-level rate limits support production operation, although some capabilities are in Beta or rolling out and availability is limited to supported Databricks regions.
Our vision for the ClickHouse Grafana plugin
ClickHouse is widely used with Grafana for observability and real-time analytics, but requiring users to write raw SQL creates onboarding and exploration friction. To simplify the user experience, planned updates to the ClickHouse Grafana plugin introduce native interfaces such as click-to-filter actions, categorized OpenTelemetry attribute displays, and guided variable editors. A new compact search mode leverages the hasToken function for fast log filtering, while a visual metrics builder simplifies aggregating OpenTelemetry and Kubernetes metrics without manual queries. Prebuilt dashboards for logs and traces will accelerate initial setup alongside automated annotations derived from service version changes and Kubernetes lifecycle events. Future prototypes also explore bidirectional SQL AST parsing for seamless query editing, JWT user identity forwarding for fine-grained access control, and conversational query generation.
Inside VAKRA: Reasoning, Tool Use, and Failure Modes of Agents
VAKRA is an executable, tool-grounded benchmark for evaluating whether AI agents can complete enterprise-like workflows that combine API calls, document retrieval, and natural-language constraints. It provides more than 8,000 locally hosted APIs backed by real databases across 62 domains, plus domain-aligned documents, and uses full execution traces to assess multi-step behavior. The four capability groups cover API chaining, dashboard-API tool selection, multi-hop dashboard reasoning, and multi-hop multi-source reasoning with multi-turn dialogue and optional tool-use policies; tasks can require three to seven reasoning steps, while some segments specify one to twelve tool calls or up to five logical hops. Reported analyses find that models perform best on single-hop questions, degrade as API hops increase, and often violate policies or fail to retrieve enough information when constraints restrict the relevant source, exposing a gap between isolated tool competence and end-to-end reliability.
Amplitude ships 3x more production code with Cursor
Amplitude’s engineering team built toward an autonomous development pipeline that moves software from idea to production with minimal developer intervention, using Cursor to connect customer feedback, observability, code reviews, and other lifecycle context to agents. Cloud agents run in isolated, scalable VMs with complete development environments, enabling parallel, long-running, self-testing work that local machines could not support at the same scale. Automations now turn Slack bug reports and feature requests into Linear tickets and pull requests, perform hourly legacy migrations, and classify pull-request risk, while Bugbot reviews and autofixes low-risk changes. Since adopting cloud agents, Amplitude reports a threefold increase in weekly production commits, more than 1,000 unprompted agent runs per week, and Cursor as a top-three contributor by commit volume. About 60–70% of low-risk pull requests merge without additional developer work, while high-risk changes are routed to appropriate reviewers; the next goal is automating CI/CD, build validation, and deployment.
Better AI models enable more ambitious work
An eight-month study of developers at 500 companies using Cursor examined whether improved AI mainly increases existing work or enables tasks previously out of reach. The research, conducted with Professor Suproteem Sarkar of the University of Chicago Booth School of Business, covered releases of Opus 4.5 and GPT-5.2, finding average weekly messages per user rose 44%. Developers initially used better models for more work of similar complexity, then after a 4–6-week lag shifted toward harder tasks: low-complexity messages grew 22%, while high-complexity messages grew 68%, mostly in the final six weeks. Growth varied by sector, led by media and advertising at 54%, software and developer tools at 47%, and finance and fintech at 45%. Documentation, architecture, code review, and learning grew more than UI and styling, suggesting expanding AI-generated codebases increase demand for management, understanding, and review while opening broader productive opportunities.
Meet HoloTab by HCompany. Your AI browser companion.
HoloTab is a Chrome extension that brings HCompany’s computer-use AI into the browser. It navigates websites, fills fields, and makes decisions from a user’s description, with vision models, action planning, and interface understanding operating underneath without setup or technical skills. The announcement follows the March 31st release of Holo3, described as the company’s most advanced computer-use model to date. For repetitive workflows, users can record their actions and narration in a tab; HoloTab turns the recording into a routine that can later be rerun or scheduled. The stated goal is to make computer-use AI accessible beyond teams with engineers who can connect such systems, and the extension is presented as free and available for everyone today.
Building with Modal and the OpenAI Agents SDK
The post walks through constructing a custom agent harness with the OpenAI Agents SDK and Modal Sandboxes, starting from a minimal coding agent that executes shell commands directly on the host. To improve safety, the agent is moved into isolated Modal Sandboxes, which are secure Linux environments that can be equipped with GPUs for compute-intensive tasks like training models. Memory is added via Sessions to accumulate context across turns, and subagents are introduced to handle focused subtasks with fresh context windows, preventing context bloat in the orchestrator while allowing long-horizon work. An async subagent pool enables parallel execution, with quotas to limit GPU spend, and filesystem snapshots allow checkpointing and deduplication of setup work across subagents, saving time and offloading memory. The harness is applied to the Parameter Golf challenge, where an orchestrator delegates parallel training experiments across different ML frameworks on GPUs, demonstrating autonomous research capabilities with composable skills.
Resend CLI 2.0
Resend CLI 2.0 gives terminal-based access to the full Resend API and adds four features: sending emails from local React (.tsx/.jsx) files, managing automations and events, one-command webhook listening with automatic tunneling, and Agent Skills for autonomous API interaction. Users can now create automation runs, inspect results, and build multi-step workflows directly from the command line. The webhook listener automatically sets up a tunnel and forwards all event types to a local server, eliminating the need for separate tunneling tools. Agent Skills provide a structured interface enabling AI agents to independently manage their own inbox, sign up for accounts, receive confirmation emails, process attachments, and respond without manual intervention. Other improvements include programmatic log retrieval, enhanced API key storage via the system keychain (macOS Keychain, Linux Secret Storage, Windows Credential Manager), and the deprecation of the team command in favor of auth.
Interact with agent-created visualizations in canvases
Cursor introduces canvases, allowing agents to represent information through interactive visual interfaces instead of chat text or markdown. In the Agents Window, these durable artifacts sit alongside tools including the terminal, browser, and source control, and agents can build dashboards, PR review interfaces, architecture diagrams, and eval-failure investigations. Canvases use a React-based UI library with first-party tables, boxes, diagrams, and charts, plus existing Cursor components such as diffs and to-do lists; skills can teach agents to create specific canvas types. Cursor reports that combining observability data from Datadog, Databricks, and Sentry MCPs with local debug files makes time-series analysis easier, while eval skills group rollout failures to expose hidden harness bugs. The company is also using canvases to visualize autoresearch progress and current hypotheses during client-performance experiments.
Autoresearch isn’t just for training models
Repeated visual-regression failures made Polaris CI feedback cycles take about 30 minutes, prompting an attempt to optimize build time rather than fix individual tasks. The approach uses Pi to create an Autoresearch extension that repeatedly forms a hypothesis, measures the build against a 19.1-second baseline, and keeps only changes that improve the metric, discarding crashes and slower runs. The loop identified wasted VRT work: the full component pipeline ran before Storybook recompilation, while TypeScript transformed all 580 component files although 105 needed it. Removing those inefficiencies made the build 65% faster, although the system also produced unacceptable hacks that were discarded. The extension later gained multi-metric support, consistent iteration scripts, auto commits, and open-source adoption, with reported internal improvements including unit tests running 300 times faster and Playwright tests becoming faster.
Index-based pruning in ClickHouse
ClickHouse accelerates analytical queries by minimizing the amount of data read from disk across granules. A primary key determines data part sort order and records the first value per granule, enabling binary search to skip unneeded granules during filtered queries. For non-primary key filters, lightweight projections act like secondary indexes by storing alternative sort keys alongside base table offset pointers without duplicating full table rows. Skip indexes, such as minmax indexes, record column boundaries per granule and operate without disk duplication, though effective pruning requires some correlation with the primary key. Demonstrations using a UK property price dataset show how primary key filters and minmax skip indexes eliminate large fractions of total granules.
New Research Reveals Overcoming Legacy Tech Issues Key to AI Success
An IDC survey of 1,400 organizations across eight Asia/Pacific markets reveals that legacy architectures and accumulated data debt severely hinder enterprise artificial intelligence initiatives. While 43% of surveyed organizations cite legacy systems as a major obstacle to application development, 95% report project delays and 90% have experienced failed modernization attempts. In contrast, high-performing organizations treat modernization as a continuous discipline across people, processes, and cloud-centric data platforms capable of processing both structured and unstructured workloads. These leading enterprises achieve nearly three times more digital revenue growth than peers still constrained by technical debt. IDC notes that addressing technical and data debt proactively avoids escalating AI failure rates and delivers superior return on investment compared to superficial lift-and-shift migrations.
What's new in ClickStack - March 2026
ClickStack announced several observability updates in its March 2026 release to improve root cause analysis, dashboard organization, and local development. The Event Deltas tool now displays baseline span distributions immediately, uses deterministic cityHash64 sampling for trace heatmaps, and applies proportional comparison scoring to accurately evaluate attributes across uneven sample sizes. Furthermore, users can directly include or exclude values from attribute comparison bars to streamline query creation during incident investigations. Dashboard capabilities expanded with a curated OpenTelemetry template gallery, multi-value filtering with optional lookup WHERE clauses, and broadened metric display units. Finally, local mode introduces browser localStorage persistence through createEntityStore, allowing developers to save dashboards and favorites without a running backend API.
Agentic reasoning in practice: Making sense of structured and unstructured data
Enterprise questions that connect structured records with unstructured text are difficult for single-turn retrieval systems to answer reliably. The post evaluates Databricks Agent Bricks Supervisor Agent (SA), a declarative agent builder built on the internal aroll framework, across STaRK’s Amazon, MAG, and Prime domains and six KARLBench tasks. SA decomposes questions, routes sub-questions to Genie or Knowledge Assistant, combines results, and can reformulate queries or use SQL JOINs when initial retrieval is incomplete. It outperforms rerun SoTA baselines by 4% on STaRK-Amazon, 21% on MAG, and 38% on Prime, while delivering a 23% improvement on FinanceBench and consistent gains across KARLBench; the authors attribute the results to multi-step tool selection, reflection, and precise instructions rather than custom code.
Top topics
- Architecture834
- Performance816
- LLMs660
- Machine Learning622
- Open Source563
- AI Agents553
- AI470
- Developer Experience448
Weekly digest
The best engineering reads, once a week.