---
title: "Latest reads"
description: "The engineering internet, summarised so you can actually read it."
---

# Latest reads
> The engineering internet, summarised so you can actually read it.

## Articles

### [Securely send first-party conversion signals with Snapchat Conversions API on Databricks Marketplace](https://yomu.fyi/post/securely-send-first-party-conversion-signals-with-snapchat-conversions.md)
- Company: [Databricks](https://yomu.fyi/company/databricks.md)
- Author: Nikhil Gaekwad, Jen McNamee, Katy Yuan
- Published: Apr 28, 2026
- Type: Announcement

Snapchat Conversions API is now available as a Databricks Marketplace listing that lets teams activate first-party conversion signals directly from governed Lakehouse data. The included pre-built notebook reads gold-layer tables, batches web, app, and offline events, and supports deduplication with Snap Pixel, allowing server-side signals without double-counting. Because it runs in the Databricks environment, Unity Catalog preserves governance, lineage, security, and access controls while teams control what data is shared and when. The post says fuller server-side coverage can raise Event Match Quality, while citing that advertisers who improved their EQS from “Poor” to “Good” saw 26% higher ROAS and 49% lower CPI. It also describes fresher custom audiences and Lookalikes plus more complete deduplicated measurement.


### [Connecting Agents to Decisions](https://yomu.fyi/post/connecting-agents-to-decisions.md)
- Company: [Palantir](https://yomu.fyi/company/palantir.md)
- Author: Palantir
- Published: Apr 28, 2026
- Type: Explainer

Palantir’s Ontology is presented as a decision-centric architecture for human-agent operations, intended to represent how enterprises make and execute decisions rather than only storing or analyzing data. It combines data, logic, action, and security in a shared, dynamic representation that can connect operational data, computational processes, workflow execution, and policy controls. The data layer includes structured, streaming, edge, unstructured, imagery, and decision data, with decision lineage recording the data version, application, and decision context; logic binding connects heterogeneous business, modeling, planning, and optimization assets for human and agent workflows. The architecture exposes approved actions to agents with configurable human review and guardrails, while capturing decisions as learning material for improving workflows and agentic memory. It concludes that a real-time ontology can anchor AI in enterprise conditions while controlling where automation is allowed.


### [Introducing the OSSCAR Index](https://yomu.fyi/post/introducing-the-osscar-index.md)
- Company: [Supabase](https://yomu.fyi/company/supabase.md)
- Author: prashant
- Published: Apr 28, 2026
- Type: Announcement

Supabase has launched the OSSCAR Index, a quarterly ranking of fastest-growing open source organizations, with the site, data, and scoring code released as open source. It addresses the limits of rankings based on accumulated stars, downloads, or contributors by measuring community growth within two divisions: Emerging organizations below 1,000 GitHub stars at quarter-start and Scaling organizations at or above that threshold. Scores combine normalized net new stars, unique contributors, and npm, PyPI, and Cargo downloads using an L² norm, with minimum thresholds and no penalty for missing or declining signals. The Q1 2026 edition identifies Openclaw as a breakout project, places AI-agent projects prominently in Emerging, and reports global representation; quarterly updates, expanded package coverage, an RFC process, and historical rankings are planned.


### [Formatting an entire 25 million line codebase overnight: the rubyfmt story](https://yomu.fyi/post/formatting-an-entire-25-million-line-codebase-overnight-the-rubyfmt-st.md)
- Company: [Stripe](https://yomu.fyi/company/stripe.md)
- Author: Fable Tales
- Published: Apr 28, 2026
- Type: Problem & solution

Stripe’s 25-million-line Ruby codebase needed a true autoformatter; before rubyfmt, no industry-wide Ruby formatter existed, and earlier tools had crashed on its files. rubyfmt began as a zero-configuration open-source project motivated by formatting disputes, Ruby’s slow startup, and the uncertainty faced by engineers arriving from other languages. To meet a 100ms target, it initially used a hand-authored C launcher and later moved formatting into Rust, with ripper parse trees first passing through JSON and eventually being deserialized directly from Ruby VALUE objects. Stripe adopted per-file opt-in, expanded coverage cautiously, and formatted the codebase on a Saturday in 2024; the excerpt reports that all 42-million-line Ruby codebase is now formatted. In 2025, migrating from ripper to Prism removed the need to link a Ruby VM and made rubyfmt smaller and dramatically faster.


### [How leading tech companies are killing the builder's tax with Lakebase](https://yomu.fyi/post/how-leading-tech-companies-are-killing-the-builder-s-tax-with-lakebase.md)
- Company: [Databricks](https://yomu.fyi/company/databricks.md)
- Author: Amey Banarse, Madelyn Mullen
- Published: Apr 27, 2026
- Type: Explainer

The post argues that AI-native applications face a “builder’s tax” because operational databases and lakehouse analytics are separated by CDC, ETL/ELT, and reverse ETL pipelines. This architecture can leave AI systems with stale data while duplicating governance and shifting engineering effort toward orchestration and failure management. It presents Lakebase, a fully managed serverless Postgres engine integrated with the Databricks Data + AI Platform, as a shared operational foundation where applications, agents, analytics, and governance use the same data. The proposed pattern combines a lakehouse intelligence layer, a low-latency Lakebase execution layer, and a continuous learning loop, with pgvector supporting agent memory and semantic search. Reported customer examples include faster feature onboarding, reduced on-call disruption, higher agent throughput, and consolidated data estates, though the results vary by organization.


### [Putting Fairness in the Driver’s Seat](https://yomu.fyi/post/putting-fairness-in-the-driver-s-seat.md)
- Company: [Uber](https://yomu.fyi/company/uber.md)
- Published: Apr 27, 2026
- Type: Announcement

Uber describes its deactivation principles and safeguards for drivers and couriers, motivated by the stress of losing app access and the need for decisions that are fair, accurate, and transparent. Most access issues are temporary, while permanent deactivations are described as rare and intended to protect platform safety; users should receive notice, relevant details, and an opportunity to respond whenever possible. The updated process lets drivers view eligible reports, submit context including Record My Ride audio or video, and have reports reviewed alongside their information, while systems filter suspected fraudulent or false claims. The Review Center explains deactivation reasons and appeal options, supports uploads, and sends every appeal to a real person, with additional measures including excluding low ratings caused by factors outside a driver's control, limiting trip categories when appropriate, and covering third-party testing for disputed impairment claims.


### [From Clicks to Conversions: Architecting Shopping Conversion Candidate Generation at Pinterest](https://yomu.fyi/post/from-clicks-to-conversions-architecting-shopping-conversion-candidate.md)
- Company: [Pinterest](https://yomu.fyi/company/pinterest.md)
- Author: Pinterest Engineering
- Published: Apr 27, 2026
- Type: Problem & solution

Pinterest built a dedicated shopping candidate generation model to optimize for lower-funnel purchase conversions rather than relying strictly on engagement retrieval models. Because offsite conversion events are sparse and delayed, engineers trained a single multi-surface two-tower model using conversion data supplemented by duration-weighted click engagement and unengaged ad impressions as hard negatives. The two-tower retrieval architecture incorporates a parallel DCN v2 and MLP layer structure that allows simultaneous learning from the same inputs without information bottlenecks. Later revisions transitioned the system from a multi-head loss structure to a unified single-head multi-task architecture paired with an advertiser-level objective. Deployed across more than 600 million monthly active users, the system increased shopping conversion volume by 2.3% and boosted return on ad spend by 3.1% for United States shopping campaigns.


### [Powering KPMG UK Audit's AI future with Databricks](https://yomu.fyi/post/powering-kpmg-uk-audit-s-ai-future-with-databricks.md)
- Company: [Databricks](https://yomu.fyi/company/databricks.md)
- Author: Mark Wallington, Greta Nasai
- Published: Apr 27, 2026
- Type: Problem & solution

KPMG UK is evolving its Audit data platform with Databricks to support AI-enabled analytics while preserving governance, rigour, and professional standards. The program converges structured data, advanced analytics, and AI on a unified cloud-native Lakehouse, with Databricks SQL replacing core SQL Server workloads and Delta underpinning the platform. Lakebridge assessed migration complexity, while Databricks-hosted large language models including Claude Sonnet and Genie Code helped convert T-SQL, refactor stored procedures, modularize queries, and suggest Delta-based optimizations under engineer review. This reduced refactoring time by around 60% and enabled modernization of more than 400 scripts and stored procedures in roughly three months. Databricks SQL Serverless provides elastic compute for spiky workloads, while Genie offers traceable, version-controlled SQL and Delta Sharing supports governed data exchange.


### [For Teams – Tools & craft – Notion](https://yomu.fyi/post/for-teams-tools-craft-notion.md)
- Company: [Notion](https://yomu.fyi/company/notion.md)
- Published: Apr 27, 2026
- Type: Explainer

The page presents a collection of Notion content about teams, tools, workplace practices, and the use of AI in organizational work. Its entries cover Custom Agents for automating repetitive busywork, teamwork skills, launch orchestration, and the changing role of team size. Several pieces focus on disconnected company knowledge, describing Notion AI as a way to bring information together and turn knowledge into action, while other entries examine AI operating systems and workflows. The page also includes survey-based perspectives from Notion, including research involving more than 3,000 global knowledge workers, 85-plus interviews, and 650 European business leaders, alongside customer stories from companies such as Merge, Ramp, Chronicle, Toyota, Affirm, and Mercury.


### [ClickHouse at Google Cloud Next '26](https://yomu.fyi/post/clickhouse-at-google-cloud-next-26.md)
- Company: [Clickhouse](https://yomu.fyi/company/clickhouse.md)
- Author: ClickHouse
- Published: Apr 27, 2026
- Type: Announcement

ClickHouse announced several product updates and integrations designed to deepen native compatibility with Google Cloud. Migration of ClickHouse Cloud to Google Axion ARM-based processors demonstrated a 30 to 55 percent query speed increase and reduced data load times from 64 to 26 seconds on ClickBench benchmarks. An integration with Google Lakehouse Runtime Catalog enables querying Iceberg tables on Google Cloud Storage via the DataLakeCatalog engine without data movement. Furthermore, Bring Your Own Cloud reached general availability on Google Cloud Marketplace, and ClickPipes added an unordered mode using Pub/Sub notifications for GCS ingestion. Google Antigravity integrated ClickHouse through an MCP server to translate natural language into SQL and generate interactive React charts.


### [Inside one of the first production deployments of Lakebase: LangGuard's agentic workflow governance engine](https://yomu.fyi/post/inside-one-of-the-first-production-deployments-of-lakebase-langguard-s.md)
- Company: [Databricks](https://yomu.fyi/company/databricks.md)
- Author: Venkat Raghavan, Jason Keirstead, Ravi Srinivasan, Nina Williams, Amelia Westberg
- Published: Apr 27, 2026
- Type: Explainer

LangGuard is presented as a runtime enforcement layer for enterprise agentic workflows, monitoring actions, decisions, tools, credentials, and intent across connected systems. Its GRAIL data fabric records multidimensional trace data, builds a live knowledge graph, and evaluates allow/deny/modify decisions against policy before tools, datasets, or models are accessed. The deployment uses Databricks Lakebase as the operational system of record, relying on PostgreSQL, serverless autoscaling, scale-to-zero, compute-storage disaggregation, hot-data caching, and copy-on-write branching for policy testing. LangGuard chose this architecture to handle bursty trace writes and low-latency enforcement reads without provisioning for peak demand, while keeping operational data available to Databricks analytics and AI capabilities without additional ETL. The stated next step is predictive governance: training behavioral models on historical traces to flag anomalous agent behavior before a policy violation.


### [How to build scalable web apps with OpenAI's Privacy Filter](https://yomu.fyi/post/how-to-build-scalable-web-apps-with-openai-s-privacy-filter.md)
- Company: [Hugging Face](https://yomu.fyi/company/hugging-face.md)
- Author: yuvraj sharma, Freddy Boulton, Abubakar Abid
- Published: Apr 27, 2026
- Type: Tutorial

The article presents three web applications built around OpenAI’s open-source Privacy Filter, a PII detector that labels eight categories in one forward pass over a 128k-token context. Document Privacy Explorer extracts PDF or DOCX text and returns category-labeled spans, while Image Anonymizer combines Tesseract word boxes with character offsets to produce editable pixel redaction rectangles. SmartRedact Paste replaces detected spans with category placeholders, stores public and token-gated views, and evicts expired pastes in a daemon thread. Across the apps, gradio.Server’s @server.api routes model work through Gradio’s queue and gradio\_client, while plain FastAPI routes serve custom HTML, views, and lightweight lookups. The implementations keep browser-side filtering, canvas edits, and PNG export out of server round-trips, and the paste service is about 200 lines including storage.


### [The next generation of Databricks Genie](https://yomu.fyi/post/the-next-generation-of-databricks-genie.md)
- Company: [Databricks](https://yomu.fyi/company/databricks.md)
- Author: Ken Wong, Dillon Morrison, Richard Tomlinson
- Published: Apr 26, 2026
- Type: Announcement

The next generation of Databricks Genie expands the assistant beyond individual Genie Spaces, combining structured and unstructured enterprise data in a unified chat experience. It reuses logic from certified Genie Spaces, governed dashboards, and Databricks Apps, with metadata-based routing that prioritizes higher-trust sources, while new reasoning models and agent architecture handle questions spanning multiple domains. Built-in connectors for Google Drive and SharePoint, plus MCP support, let Genie access knowledge stores and take actions; Unity Catalog AI Gateway manages these connections. The experience replaces Databricks One as an account-level global home, adding domains, custom URLs, unified login, automated identity management, and governance through Unity Catalog. Native iOS and Android apps extend access to dashboards, apps, and chat beyond desktop environments.


### [Model risk management in 2026: A banker's guide to the revised interagency guidance](https://yomu.fyi/post/model-risk-management-in-2026-a-banker-s-guide-to-the-revised-interage.md)
- Company: [Databricks](https://yomu.fyi/company/databricks.md)
- Author: Pavithra Rao, Jennifer Miller, Chaitanya Varanasi, Kim Hatton
- Published: Apr 25, 2026
- Type: Problem & solution

The post presents the April 17, 2026 revision to model risk management guidance as a shift toward risk-based, principles-driven oversight, replacing several earlier issuances from the Federal Reserve, FDIC, and OCC. It identifies five operational changes: materiality tiering, end-to-end lifecycle governance, reproducible effective challenge, continuous drift and performance monitoring, and application of MRM principles to GenAI and agentic systems. Its proposed response is a Databricks reference architecture built on Unity Catalog, Delta Lake, Lakeflow Declarative Pipelines, Feature Store, MLflow, Model Registry, Model Serving, and assurance tools. The design makes tiering metadata-driven and turns lineage, validation, approvals, monitoring, documentation, and retirement records into evidence generated during normal model work. The stated conclusion is that a unified substrate can reduce integration and manual evidence work, allowing regulatory changes to become configuration exercises rather than multi-quarter programs.


### [OpenAI GPT-5.5 + Codex, now available and fully governed on Databricks](https://yomu.fyi/post/openai-gpt-5-5-codex-now-available-and-fully-governed-on-databricks.md)
- Company: [Databricks](https://yomu.fyi/company/databricks.md)
- Author: Patrick Wendell, Hanlin Tang, Ahmed Bilal, Arnav Singhvi, Ivan Zhou, Harish Gaur
- Published: Apr 24, 2026
- Type: Announcement

Databricks announces native support for OpenAI’s GPT-5.5, making it available for coding workflows with Codex, enterprise agents, document pipelines, and data-driven employee workflows. Access is governed through Unity AI Gateway, which provides permissions and rate limits, configurable guardrails for PII, prompt injection, and content safety, MCP tool-call auditing, failover, and request-level observability for model and Codex interactions. The announcement describes GPT-5.5 use with Genie for natural-language analytics, Agent Bricks Custom Agents for multi-step workflows, and Lakeflow Spark Declarative Pipelines for document-focused GenAI ETL involving summarization, extraction, and classification. GPT-5.5 is available on Databricks across AWS, Azure, and GCP, with entry points including AI Playground, governed Codex endpoints, Agent Bricks, and the Foundation Model API.


### [Operational databases: How they work and when to use them](https://yomu.fyi/post/operational-databases-how-they-work-and-when-to-use-them.md)
- Company: [Databricks](https://yomu.fyi/company/databricks.md)
- Author: Databricks Staff
- Published: Apr 24, 2026
- Type: Explainer

Operational databases, or online transaction processing (OLTP) databases, support day-to-day operations by processing real-time CRUD transactions with low latency, high concurrency, and ACID guarantees. The guide contrasts them with data warehouses and OLAP systems, which favor historical analysis, complex aggregations, and read-heavy workloads, while describing ETL, CDC, and operational data stores in the OLTP-to-OLAP pipeline. It argues that traditional OLTP systems can struggle with real-time analytics and AI because of siloed data, structured-data constraints, rigid schemas, scaling limits, and missing governance features. Databricks Lakebase is presented as a hybrid solution using separate storage and compute, serverless Postgres, elastic scaling, branching, shared lakehouse storage, and open formats. The proposed workflow connects existing OLTP systems through CDC or streaming into Delta Lake for fresh SQL, BI, ML, and AI access.


### [DeepSeek-V4: a million-token context that agents can actually use](https://yomu.fyi/post/deepseek-v4-a-million-token-context-that-agents-can-actually-use.md)
- Company: [Hugging Face](https://yomu.fyi/company/hugging-face.md)
- Author: ben burtenshaw
- Published: Apr 24, 2026
- Type: Explainer

DeepSeek V4 is presented as an open model family aimed at long-running agentic workloads, with Pro and Flash MoE checkpoints offering 1M-token context windows. Its central architecture alternates Compressed Sparse Attention (CSA) and Heavily Compressed Attention (HCA), compressing KV entries by 4x and 128x respectively, while using FP8/BF16 storage and FP4 indexing. The post-training design preserves reasoning across user turns when tool calls are present, introduces a dedicated XML tool-call schema, and relies on DSec, a Rust sandbox platform for RL rollouts. At 1M tokens, Pro uses 27% of V3.2's single-token FLOPs and 10% of its KV cache, while Flash uses 10% and 7%; V4-Pro-Max reports 80.6% on SWE Verified and 0.59 MRCR 8-needle retrieval at 1M.


### [What's New in pg\_clickhouse - JSONB Support, SQL value functions, Streaming, and more](https://yomu.fyi/post/what-s-new-in-pg-clickhouse-jsonb-support-sql-value-functions-streamin.md)
- Company: [Clickhouse](https://yomu.fyi/company/clickhouse.md)
- Author: David Wheeler
- Published: Apr 24, 2026
- Type: Announcement

Recent releases of pg\_clickhouse, a Postgres foreign data wrapper for ClickHouse, introduce query pushdown enhancements and streaming result handling. Version 0.1.10 adds pushdown for JSONB accessor operators and extraction functions by mapping them to ClickHouse sub-column syntax and toJSONString calls outside SELECT clauses. In version 0.2.0, date and time functions such as CURRENT\_DATE, CURRENT\_TIMESTAMP, and clock\_timestamp() push down to ClickHouse while honoring the active Postgres session time zone and precision settings. To address memory pressure and out-of-memory risks from large foreign queries, the extension adds query result streaming to the HTTP driver with a default batch buffer of approximately 50MB. Container benchmarking against the NYC taxi dataset showed streaming reduced peak memory consumption from over 600 MiB to under 86 MiB.


### [Databricks partners with OpenAI on GPT-5.5](https://yomu.fyi/post/databricks-partners-with-openai-on-gpt-5-5.md)
- Company: [Databricks](https://yomu.fyi/company/databricks.md)
- Author: Hanlin Tang, Ahmed Bilal, Arnav Singhvi, Ivan Zhou, Harish Gaur
- Published: Apr 23, 2026
- Type: Announcement

Databricks announces a partnership with OpenAI around GPT-5.5, described as OpenAI’s strongest frontier model for enterprise agentic work, complex document reasoning, and long-horizon coding agents. The model powers Codex and is presented as able to research, analyze data, create documents and spreadsheets, operate software, use tools, check outputs, recover from ambiguity, and continue through multi-part tasks. Databricks evaluated it on OfficeQA, a benchmark built from 89,000 pages of U.S. Treasury Bulletins that tests document retrieval, table interpretation, and precise calculation. With retrieval handled, GPT-5.5 scored 64.66% versus GPT-5.4’s 57.14%; in the full-agent OfficeQA Pro Agent Harness, it scored 52.63% versus 36.10%, representing reported improvements of roughly 13% and a 46% reduction in errors.


### [NYC Driver Benefits](https://yomu.fyi/post/nyc-driver-benefits.md)
- Company: [Uber](https://yomu.fyi/company/uber.md)
- Published: Apr 23, 2026
- Type: Explainer

Eligible Uber drivers in New York City may receive minimum earnings and several protections established through an agreement involving Uber, the New York Attorney General, and New York City Taxi and Limousine Commission (TLC) rules. For trips beginning in the city, the TLC minimum-pay formula considers time, distance, and driver utilization, while incentives and guarantees count toward earnings but tips, customer-charged fees, and offline promotions do not. As of March 1, 2026, the listed rates are $1.283/mile and $0.681/mile, with different rates possible for wheelchair accessible vehicle (WAV) trips and out-of-city portions. Eligible drivers may also access Black Car Fund benefits, possible unemployment insurance, and paid sick leave, which accrues at one hour per 30 hours worked up to 56 hours yearly and can be redeemed for $18.58 per hour as of that date.


[Newer posts](https://yomu.fyi/posts/page/138.md) · [Older posts](https://yomu.fyi/posts/page/140.md)