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
Nostalgia, Coupons, and a State Machine: The Swiggy Mother’s Day Story
Swiggy’s Mother’s Day campaign turned childhood nostalgia into an interactive game where users scanned virtual “mom weapons” such as a jhadu, chappal, wooden spoon, and belt to unlock coupons and cashback. The experience used Rive’s runtime animations, state machines, ViewModel properties, and backend data binding to drive scan-triggered states, reward values, and a dynamically assembled share page. Design authored visual logic inside Rive while engineering connected typed values from the React application, including strings and booleans for coupon labels and reveal transitions. A debugging issue showed that the Rive artboard needed the correct ViewModel binding and that a numeric value could not substitute for a boolean state-machine input. The team concludes that, after a learning curve, Rive reduced custom animation work and let design and engineering proceed with less blocking.
Modal is a computer
Modal is a platform that runs programs of arithmetic and logic operations on information, making it a computer. It operates on hardware from dozens of cloud providers, building a virtualization layer that aggregates, isolates, and multiplexes demand for increased resource efficiency. Programs run in containers that receive inputs from the container runtime, similar to how a typical computer’s ALU fetches inputs from memory. Information in Modal lives in a tiered storage hierarchy, from container memory to object storage, with caches like page cache and ephemeral disk. Input and output pass through the Modal Input/Output Plane for functions and sandboxes, or the Routing Plane for HTTP servers, with direct Tunnels for extreme performance needs.
Welcoming Matt Gilbert, our new Data Analyst
Resend announces Matt Gilbert as a Data Analyst on its Trust & Safety team. He spent 10 years as a Deliverability Engineer at Mailchimp, handling Delivery-team data and maintaining sending infrastructure that pushed around a billion emails a day, after earlier work on the Mandrill transactional mail service. Gilbert describes email delivery as a multidisciplinary field spanning MTA operations, data engineering, analytics, mail-queue performance, and troubleshooting delivery and inbox-placement problems, which led him to specialize in data engineering and analytics. He says he joined Resend because email remains an open, globally available way for services to communicate, and plans to help build data-driven capabilities and intelligence that enable customers to send better and reach inboxes quickly.
LiteRT.js, Google's high performance Web AI Inference
Google announces LiteRT.js, a JavaScript binding of LiteRT that runs .tflite AI and ML models directly in web browsers. It targets local inference for JavaScript and TypeScript applications, with privacy benefits, zero server costs, and low latency, while providing smoother deployment for existing .tflite models. Through WebAssembly, it exposes LiteRT’s optimized runtime across XNNPACK CPU, ML Drift with WebGPU, and experimental WebNN for NPUs; LiteRT Torch supports one-step PyTorch conversion, and AI Edge Quantizer enables layer-specific quantization. The initial release includes the @litertjs/core npm package, demos, and examples covering YOLO26, depth estimation, and Real-ESRGAN upscaling. In controlled benchmarks on a 2024 Apple MacBook Pro with M4, it outperformed other web runtimes by up to 3x, while WebGPU or WebNN delivered 5–60x speedups over CPU execution; results may vary with hardware, thermal conditions, and browser drivers.
Ben Bernanke appointed to Anthropic’s Long-Term Benefit Trust
Anthropic’s Long-Term Benefit Trust (LTBT) appointed Dr. Ben Bernanke, former Federal Reserve Chair and Brookings Institution Distinguished Fellow, as its newest member. The independent body helps hold Anthropic to its mission of responsibly developing advanced AI for humanity’s long-term benefit, and its Trustees advise the board and leadership on critical decisions involving AI’s risks and societal impacts. Bernanke brings research and policy experience spanning the Great Depression, financial crises, and his 2006–2014 Federal Reserve tenure; he will contribute to Anthropic’s economic research, including work on how AI changes economies and workforces. The LTBT can appoint members to Anthropic’s board, while its Trustees hold no company equity or profit share and are compensated only for their service.
Progress from our Frontier Red Team
Anthropic’s Frontier Red Team reports on national-security risks from frontier AI models, drawing on work across four releases over the past year. It finds early-warning signs of rapid progress in dual-use capabilities: Claude reached undergraduate-level performance on some cybersecurity evaluations and exceeded expert baselines in selected biology tasks, while current models remain below thresholds for substantially elevated national-security risk. The team compares model generations on CTF and Cybench tests, realistic roughly 50-host cyber ranges, virology troubleshooting, LabBench, and controlled biosecurity studies, including expert red-teaming. Models still fail autonomously in realistic network environments and make critical mistakes in weaponization planning, although researcher-built tools enabled an attack resembling a known large-scale theft of personally identifiable information; the team is expanding automated evaluations and pursuing government partnerships and stronger safeguards.
Frontier threats red teaming for AI safety
Frontier threats red teaming applies adversarial testing to evaluate whether AI models could create national-security risks, using biology as a test project. Anthropic worked with biosecurity experts to define threat models, probe models for more than 150 hours, learn jailbreak techniques, and build quantitative, repeatable evaluations in a secure interface without public-deployment safety monitoring. The project found that frontier models can sometimes produce sophisticated, accurate, detailed biological knowledge, with capability increasing alongside model size and potentially with tool access; the authors assess unmitigated risks as possibly near-term. Training changes and classifier-based filters reduced harmful outputs or access to chained expert information, and these mitigations were deployed in the public-facing frontier model. Future work will test LLM speedups over search engines, examine newer and smaller adapted models, and expand collaboration, disclosure, and independent evaluation.
Anthropic's core views on AI safety
Anthropic argues that AI could have an impact comparable to the industrial and scientific revolutions, potentially within the coming decade, because scaling laws link increased training computation with predictable capability improvements. It says progress may continue toward systems matching or exceeding human performance across many intellectual tasks, while society faces disruption and competitive pressure to deploy systems that are not reliably helpful, honest, and harmless. The company’s safety strategy is empirically driven and multifaceted, emphasizing scalable supervision, mechanistic interpretability, process-oriented learning, and research into how systems learn and generalize. It also studies capability evaluation, red teaming, harmful behavior, economic effects, and policy implications. The stated goal is a broad portfolio of safety work that can adapt as evidence clarifies which future scenario is unfolding, including less optimistic possibilities.
Automating cross-repo documentation with GitHub Agentic Workflows
Maintaining documentation across separate repositories often leads to severe lag because technical writers must reverse-engineer shipped features weeks after release. To address this in the Aspire project, the team implemented an automated pipeline using GitHub Agentic Workflows to bridge the product and documentation repositories. When product pull requests merge, a bash step maps milestones to docs release branches before an LLM agent evaluates the diff, drafts documentation updates, and emits structured pull request intents. A dedicated safe-outputs handler materializes these drafts via a scoped GitHub App and assigns the original code reviewers to verify accuracy. Across 396 product pull requests, the system generated 82 documentation pull requests that all merged with a median turnaround time of 44.8 hours.
Navigating a Synapse Migration to Databricks
The post presents migration from Azure Synapse to Databricks as a way to reduce platform complexity while supporting data engineering, analytics, machine learning, real-time pipelines, and AI applications. It distinguishes migration paths for Dedicated SQL Pools, serverless SQL, Spark Pools, orchestration, governance, and downstream BI connectivity, noting that accumulated T-SQL logic and dependencies require the most redesign, testing, and validation. The proposed programme uses discovery and assessment with Lakebridge Profiler and Analyzer, then selects an architecture, sequences workloads, pilots a real use case, and scales migration in waves. It recommends automating code conversion, reconciling row counts, aggregations, hashes, and tolerances, avoiding direct Synapse distribution-to-Delta partition mappings, and planning compaction, enablement, and rollback. Reported examples include Casey’s halving operational data delivery times from eight hours to four and Italgas cutting workload costs by 73%.
Create a LangChain Deep Agents Harness Profile for NVIDIA Nemotron 3 Ultra to Improve Performance
The tutorial shows how to create and validate a LangChain Deep Agents harness profile for NVIDIA Nemotron 3 Ultra using an evaluation benchmark rather than fine-tuning the model. It starts with a baseline, analyzes failed trajectories, and iterates on prompts, exclusions, or added middleware while rerunning tests to detect regressions. One failure occurred when read_file returned the first page of a large file without signaling that more content existed, so the model answered before paginating to the line containing opal-fox-91. The proposed ReadFileContinuationNoticeMiddleware appends a continuation notice when a full page is returned, raising the reported benchmark score from 94/127 to 96/127 and resolving all three read_file tests. The source also describes automating this propose, verify, and retain-or-reject loop, with safeguards against overfitting and regressions.
Uber Black vs Chauffeur vs Limo Service Comparison Guide
Choosing between Uber Black, a traditional chauffeur service, and a limo depends on whether the priority is flexible premium travel, a dedicated driver, or a special-event vehicle. Uber Black combines luxury sedans and SUVs, professional drivers, upfront pricing, cashless payment, and on-demand or scheduled booking through Uber Reserve, from 15 minutes to 90 days ahead. Chauffeur services generally require advance reservations and charge hourly or flat rates for extended, customized itineraries, while limo providers commonly use deposits, hourly minimums, and specialty vehicles for weddings, proms, and group celebrations. The comparison concludes that Uber Black is often the practical choice for airport transfers, business travel, dinners, and other premium rides, whereas chauffeur and limo services fit dedicated multi-stop engagements or occasions where the vehicle is central to the event.
Introducing pg_re2, fast, RE2-powered regular expressions in Postgres
pg_re2 is a Postgres extension that brings RE2-powered regular expression functions and operators to Postgres, designed to address both execution speed and syntax compatibility with ClickHouse. It uses finite-automata-based RE2 rather than Postgres's backtracking POSIX engine, and version 0.4.0 adds the @~ operator plus btree and GIN index support. Benchmarks on an AWS m6g.2xlarge with PostgreSQL 18 report RE2's functions running at 1.8x to 8.6x the speed of native equivalents, while indexed operations are 1.1x to 1.8x faster. With pg_clickhouse, using RE2-backed functions such as re2match() keeps regex behavior consistent between Postgres execution and ClickHouse pushdown, including patterns such as Quake\b. The extension duplicates ClickHouse's suite for string search, replacement, extraction, and splitting.
Benchmarking Coding Agents on Databricks’ Multi-Million Line Codebase
Databricks built a coding-agent benchmark from reviewed tasks drawn from its multi-million-line codebase, spanning languages including Scala, Go, Rust, Python, and TypeScript. It uses recent human-written pull requests, held-out tests, standard model and harness setups, and manual validation rather than an LLM judge; Git history was sealed after traces showed agents could recover solutions. Results clustered models and harnesses into three capability tiers and showed that quality, cost per task, and token price can diverge: GLM 5.2 statistically tied Opus 4.8 on quality while costing $1.28 versus $1.94 per task, whereas Sonnet 5 cost $2.09 versus Opus’s $1.94. The same model through different harnesses produced cost differences of more than 2x in some cases while quality remained the same; Pi sent about three times less context per turn. Databricks plans to shift routine work toward cheaper models, deploy open models such as GLM as daily drivers, and improve model-and-harness selection.
Running Low-Latency Analytical Workloads with GPU-Accelerated Presto on NVIDIA GB200 NVL72
GPU-accelerated Presto integrates NVIDIA cuDF algorithms and NVLink to deliver low-latency analytical query execution on massive datasets. Benchmarks derived from TPC-H across scale factors 1K and 3K reveal that a single NVIDIA DGX B200 system with eight GPUs outperforms 8- to 10-node Intel Xeon clusters by up to 8.2x. Scaling out across multinode NVIDIA GB200 NVL72 systems paired with IBM Storage Scale achieved additional performance gains. Enabling NVIDIA GPUDirect Storage, tuning I/O parameters, adjusting thread counts, and applying query rebatching delivered a cumulative 64% reduction in query runtimes. Topology-aware GPUDirect Storage cold reads proved roughly twice as fast as standard POSIX reads by avoiding CPU memory staging and NUMA boundary crossing penalties.
Uber Advertising sets commerce media in motion at Cannes 2026
At Cannes Lions 2026, Uber Advertising presented commerce media as a way for brands to connect with consumers across real-world journeys, from discovery and ordering to mobility and post-event decisions. Its EMARKETER session with Sarah Marzano and Kristi Argyilan framed commerce media as expanding beyond traditional retail environments, while announcements including Offsite Ads, Mobility Offers, Deal Drops, Reorder Rewards, Uber Marketing Manager, and APIs addressed what advertisers can activate and how they can manage and measure it. Programming also examined AI’s effects on search, shopping, planning, recommendations, and marketing leadership, alongside sports-fan behavior and cultural moments. Across the Uber Villa and activations involving flights, airport screens, rides, dining, boats, and shuttles, the company positioned movement as the connective thread between media, commerce, and everyday experience.
How to Evaluate an Enterprise Analytics Platform
Enterprise analytics platform evaluations often overemphasize dashboard interfaces, although the larger decision concerns whether analytics, AI and agents share data, semantics and governance. The post distinguishes point solutions from a unified platform and proposes seven evaluation criteria: workload fit, architecture and openness, governance and compliance, performance and scalability, adoption and usability, AI and ML readiness, and total cost of ownership. It recommends mapping current and three-year workloads, testing production-scale data with realistic concurrency, measuring p95 latency, and examining governance, usability, contracts and operational complexity in a proof of concept. Lakehouse architecture, open formats such as Delta Lake and Apache Iceberg, and shared controls are presented as ways to reduce context gaps; Databricks is offered as a practical example using Unity Catalog, Genie and Agent Bricks. The conclusion favors a weighted, three-year assessment over a feature comparison.
Open, convenient and predictable: Introducing Provisioned Throughput
Companies adopting open-weight models for production workloads have faced a trade-off between convenient but best-effort serverless inference and powerful but complex dedicated GPU instances. Together AI’s Provisioned Throughput fills that gap with reserved inference capacity sold as Provisioned Throughput Units (PTUs). Each PTU is priced at $0.05 per minute and guarantees a fixed rate of tokens per minute for a specific model, with a 99% uptime SLA. Input, cached input, and output tokens draw down capacity at different rates, letting teams optimize for their traffic shape. Available now for MiniMax M3 and GLM-5.2 with capacity in North America and EMEA and a one-month minimum term, at full utilization costs can run roughly $0.36 per million input tokens and $2.16 per million output tokens, up to 90% below proprietary alternatives.
Sunsetting new.email
Resend is sunsetting new.email, which will remain available for template exports until August 7, 2026, before stopping new prompts and sends. Launched in February 2025, the tool let anyone create email templates with natural language, using React Email components and deliverability lessons from sending millions of emails daily; its goal was consistent, beautiful, responsive email. Resend says the core bet was right, but a separate brand and stack—repository, infrastructure, deployment, and operational conventions—made the product harder to maintain, while market feedback showed users did not need new templates constantly. The product’s ideas are moving into the Resend editor, which offers URL-based brand extraction, AI block rewriting, LLM content review, and email creation and editing through the Resend MCP server.
Bridging the Domain Gap: AI Race Coach built with Antigravity and Gemini
At Sonoma Raceway, Google Developer Experts built an AI-powered race coach to turn vehicle telemetry into split-second driving advice, extending an earlier pilot that handled basic telemetry. The system uses Antigravity for stateful orchestration and ingestion, Python and Jetpack Compose for mobile processing and display, Agent Development Kit for agent coordination, and hybrid Gemma 4 edge reasoning plus Gemini API cloud analysis. A custom USB interface connected a Pixel 10 directly to the vehicle telemetry network, providing a 10 Hz stream from hundreds of sensors, while an activated on-device TPU reached 40 tokens per second for local coaching alerts delivered through text-to-speech. The test identified a throttle application zone in Turn 2 associated with a 0.1-second gain, and the initiative is scheduled to continue at Interlagos, Brazil.
Top topics
- Architecture834
- Performance816
- LLMs660
- Machine Learning622
- Open Source563
- AI Agents553
- AI470
- Developer Experience448
Weekly digest
The best engineering reads, once a week.