Loading…
Browse by type
Problem & solution reads
Problem & solution posts on Yomu, summarised newest first.
- 925 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.
Problem & solution posts
How Meta Built Forum With AI-Native Engineering
Meta built Forum using AI-native engineering, treating an AI agent as a development participant rather than a faster typing tool. The team shifted from human-in-the-loop to human-on-the-loop, where the agent autonomously plans, implements, builds, tests, and corrects its work while the engineer sets goals and approves outcomes. Success was measured by verified, production-ready results, not code volume, with automated verification signals like build status and unit tests filtering output before review. Meta adapted the codebase for legibility, gave agents tools to inspect live view hierarchy directly instead of descriptions, and created a separate sandbox build to isolate prototype code from production. This approach helped Forum reach core feature parity with Facebook Groups in eight weeks and caught a memory issue during development that roughly halved memory use and page faults.
2026 World Cup: Breaking Records, Spiky Charts and Reliability at Scale
The 2026 FIFA World Cup produced the largest planned stress test of Meta’s infrastructure, with WhatsApp setting an all-time record of over 30 million messages per second during the final and delivering 8.4 billion messages in its peak five-minute window. Traffic analysis revealed that surges in sharing primarily occurred during breaks and final whistles, not during active gameplay, shaping capacity decisions. Instagram and Facebook saw live video watch time spike 90% and 242% respectively, while recommendation traffic rose 80% week-over-week. To absorb these spikes, Meta layered pre-provisioned capacity buffers, country-level graceful degradation, cascade prevention, and early warning signals across all serving tiers. Notably, the team did not shift load between regions or freeze deployments, relying instead on year-round reliability enhancements and phased runbook execution.
How FactSet cut model training and deployment from days to hours with Databricks AI Runtime
FactSet’s Text-to-Formula (TTF) system converts natural-language investment queries from professionals into correct FactSet Query Language (FQL) formulas. Initially built on retrieval-augmented generation (RAG), the approach hit a quality ceiling as queries became more specialized and formulas more complex. FactSet adopted fine-tuned open models within a Compound AI architecture, reducing end-to-end latency by about 60 percent and improving quality for complex formulas. Fine-tuning runs on Databricks AI Runtime with serverless GPU infrastructure, where engineers choose compute based on prompt length and output needs, and the pipeline integrates with MLflow, the Lakehouse, and Lakeflow Jobs for a unified workflow. This setup cut model setup, training, and deployment from days to hours, enabling the team to retrain every two to three weeks as new data and models become available.
Why your platform team shouldn't build an AI context layer (even though they probably could)
Many engineering teams mistakenly think building an internal AI knowledge context layer is as simple as standing up a vector database and wiring a RAG pipeline to documentation. In reality, a production-grade pipeline requires five continuously maintained stages—capture, deliver, govern, organize, and validate—where each stage’s long-term cost compounds as vendor APIs change and content ages. The critical innovation is the knowledge node, a living record carrying source, ownership, and an automatically computed trust score based on authorship, usage, recency, and conflicts. Trust decays over time, prompting owners for lightweight revalidation when confidence drops, and knowledge is automatically scoped to the correct audience without manual filing. For product companies not in the knowledge-infrastructure business, the compounding maintenance burden makes building this pipeline a distraction compared to leveraging existing platforms like Stack Internal.
How Dynata scaled AI across the development lifecycle with AI-DLC and Kiro IDE
Dynata, a large first-party data company, sought to accelerate delivery across distributed teams while improving cross-functional alignment. They adopted a staged AI approach: first AI-assisted coding via Amazon Q Developer, then connecting AI to project context from Jira, Confluence, and Lucid, and finally implementing the AI-Driven Development Lifecycle (AI-DLC) methodology using Kiro IDE. AI-DLC, powered by Amazon Bedrock, structures delivery into Inception, Design, and Construction phases with Specs, steering files, and hooks. This process forced all teams to agree on requirements and design during a Day 1 workshop, surfacing gaps before coding began. Consequently, features that once took 3–4 sprints were completed in about one sprint, cutting development phase time by roughly 70% and reducing post-launch defects.
OLTP is expensive, you can ClickHouse it.
PostgreSQL’s ACID guarantees cause substantial write amplification: each write flushes WAL for durability, copies rows via MVCC for isolation, and updates every affected index, amplifying 100 bytes of row change into tens of kilobytes of I/O. Costs scale linearly with data size, and even sharded Postgres offers only linear scaling with added coordination overhead, so growing datasets drive rising bills. The post proposes offloading settled, rarely-changing data to ClickHouse, whose columnar compression shrinks datasets 10-30x and whose cloud offering costs about 5x less than comparable Postgres infrastructure while also reducing Postgres compute load through fewer indexes. ClickHouse builds indexes in background batches and runs aggregations at roughly 100x lower cost than Postgres. ClickHouse Cloud provides WalShadow for millisecond-latency CDC and pg_clickhouse FDW, enabling an application pattern that cuts the cost of serving settled data by an order of magnitude.
Selection-Based Structured Reasoning: Toward Efficient Multimodal Search Agents
Multimodal search agents typically rely on freeform reasoning before acting, but small models often produce lengthy reasoning that offers little guidance and increases inference cost. The authors propose Selection-Based Structured Reasoning (SSR), which reframes reasoning as selecting from pre-specified natural-language candidates rather than open-ended generation. SSR enables parallel scoring by using teacher-forced prefilling and a shared context KV cache to compute token likelihoods across candidates simultaneously. Evaluated on seven multimodal search benchmarks with 2B and 4B models, SSR maintains task performance under various reinforcement learning objectives and supervised fine-tuning. The framework achieves search success rates competitive with leading agents of the same scale while reducing per-turn reasoning latency by over 90% and total per-question inference latency by 28–54%.
Evaluating AI agents: moving past the vibe check
Manual vibe checks for AI agents miss silent failures like retry loops, token waste, and hallucinated parameters while plausible final answers mask them. The post outlines a structured evaluation pipeline using the open-source agent-eval toolkit, starting with critical user journeys (CUJs) that define success in concrete, testable clauses. Synthetic user simulations (via ADK) generate multi-turn scenarios for initial trace collection, which developers grade with a ‘red pen’ before automating checks. Four layers of verification are recommended: unit tests, deterministic metrics, LLM-as-a-judge with boolean rubrics (avoiding unreliable 1-to-5 scales), and integration tests, all standardized on OpenInference/OpenTelemetry traces. The rollout progresses from Day 0 manual review through Day 1 automated checks to Day 2 CI/CD gates and shared dashboards, with real production traces fed back into the golden dataset.
Italian Enterprises Leading in AI
AI models depend on flexible, high-performance data infrastructure. Leading Italian enterprises including TeamSystem, altermAInd, Brembo Solutions, and Eni are using MongoDB Atlas to power their AI initiatives. TeamSystem built a centralized AI platform with hybrid search combining full-text and vector search that delivers precise, context-aware results while minimizing hallucinations, scaling from 50,000 to 500,000 users and cutting task times by 30–40%. altermAInd solved GDPR and EU AI Act compliance at the database layer, enabling secure innovation for over 200,000 customers. Brembo Solutions uses ALCHEMIX to slash R&D formulation from months to minutes by leveraging MongoDB’s document model for diverse material data. Eni uses Atlas Search and vector search to instantly analyze years of geoscience data and Atlas Charts for real-time interactive visualizations without hand-coded dashboards. These companies show that a robust, secure, and scalable data platform is foundational to enterprise AI success.
Why AI at Work Needs a Rethink
A survey of over 6,000 knowledge workers across the UK, France, and Germany, commissioned by Notion, reveals that daily AI users report significantly more benefits than occasional users. For example, 59% of daily users in the UK say AI agents are changing team collaboration, versus 37% overall, and 48% say AI helps them demonstrate more value, nearly double the overall 25%; similar patterns hold in France and Germany. Gen Z workers are especially likely to use AI to expand their responsibilities, but they also express higher concern about job security, notably in Germany where concern was 21 points above the national average. However, AI use remains largely individual: only about 20% of respondents embed AI agents in team-wide workflows. The post argues that the next step is rebuilding AI around shared workflows and company context, as demonstrated by companies like Pennylane and Brainlabs.
From 2,500 requests to 24: Rebuilding star history without the stargazers
The post explains GitHub’s redesign of the stargazers API to provide star history without exposing user identities. The original endpoint returned paginated lists of stargazers, which tools used to reconstruct growth curves by fetching all pages, leading to privacy risks and expensive database queries that counted all stars on each request due to a pagination bug. GitHub first restricted access to stargazer lists to repository collaborators, but this broke tools that needed star history. Collaborating with maintainers and the Star History project, GitHub built a new endpoint that returns up to 30 weeks of daily star counts in a single request with no identities attached. The new endpoint reduces database work by filtering stars by time before aggregation, cutting requests for repositories like react/react from about 2,511 to 24.
How Ridgeline unified Iceberg and ClickHouse into one analytics layer for investment management
Ridgeline, a cloud-native investment management platform, rebuilt its performance analytics layer after Postgres could not handle the scale of larger customers. The team evaluated Redshift, SingleStore, Trino over Iceberg, and a custom in-memory engine before choosing ClickHouse for its analytical performance and lower cost. They adopted ClickHouse Cloud on AWS, using separate organizations for staging and production and per-customer services, with assistance from ClickHouse engineers. Historical data lands in Apache Iceberg via a Spark pipeline and is queried through ClickHouse, while recent data is written directly into ClickHouse. The architecture enabled Ridgeline to serve customers 10x larger than before and more than double its live customer count.
Optimizing Agent Reliability and Tool Calling Efficiency for Chip Design Using NVIDIA NeMo Gym
AI agents in chip design often produce plausible but incorrect artifacts, making reliability a key barrier to production adoption. NVIDIA NeMo Gym, an open library for evaluating models and agents using environments, solves this by diagnosing failure modes in multistage workflows through verification of final outputs against reference artifacts. Applied to the Siemens Fuse EDA AI Agent for standard-cell characterization, NeMo Gym uses a verifier that checks coverage, validity, relative validation, integrity, and accuracy of generated Liberty files against known references. After refining the agent’s skills and MCP tool descriptions based on trajectory analysis of failures, the composite benchmark score improved from approximately 55% to 95%, while agent turns fell from 155 to 53 and output tokens dropped from ~81K to 29K. This method generalizes to other engineering domains by emphasizing evaluation of final artifacts and domain-specific verifiers.
Inside Minerva: building scalable observability for eBay’s monetization platform
eBay’s Monetization Team lacked a dedicated observability platform, causing fee behavior issues to often be discovered through customer complaints or post-transaction analysis. Data warehouse data arrived with delays, and engineers and business teams had to manually query multiple sources. Minerva solves this with an asynchronous observability pipeline: the fee-processing service places a small record on a local queue, a background producer publishes it to an event stream, and a consumer enriches the data before storing it in ClickHouse. ClickHouse’s columnar storage and MergeTree engine allow fast aggregation queries on records with nearly 250 columns; a proof of concept on 1.1 billion rows returned queries in 0.11 to 0.72 seconds. The design keeps observability off the critical fee path, prioritizing fee-processing availability while providing configurable, context-aware monitoring.
Design Email for Humans and Agents
Apple Mail and Gmail now generate AI summaries of emails, while agents read messages to automate tasks like logging receipts or tracking deadlines. Summarizers focus on the first 100 to 150 characters, so the main point and call to action must lead the email. Semantic HTML with h1 headings, a descriptive title, lang attributes, and role=‘presentation’ on tables creates a machine-readable outline, yet the Email Markup Consortium found 74% of emails lack an h1 and 96% lack lang. Hidden text styled to be invisible can hijack summaries, as Forcepoint demonstrated, and similar attacks have targeted Gemini in Google Workspace; such practices also trigger security flags. Testing in Gmail, Apple Mail, and Outlook, along with verifying the plain text version, confirms that both humans and agents interpret the email correctly.
Rewinding the exponential: How we keep hardware usage down
Cloudflare’s Data team manages a pipeline handling over two billion events per second from tens of thousands of servers, powering Logs, SQL API, and observability. To counter exponential traffic growth against linear hardware expansion, they implemented several optimizations over five years, deferring two and a half years of hardware expansion. In 2022, they replaced a monolithic schema with product-specific schemas, reducing backbone bandwidth by 35-40%. In 2024, they switched compression from LZ4 to zstd, achieving a 40% reduction in compressed traffic with negligible CPU impact. In 2026, they decoupled pipeline transit from database redundancy via a shared ingest pool and late replication, applied sampling downstream, and introduced aggregations, eliminating about 70 GB/s of compressed traffic and halving projected 2031 data volume.
Stop focusing on the next big thing - The Intercom Blog
The post criticizes the tendency of product people to fetishize emerging trends like AI, VR, and AR without stepping back and asking why, latching onto hype rather than focus. It argues that this ‘Next Big Thing-ism’ is rooted in short-term jostling and FOMO, not focus, leading to products that seek quick exits, as seen at CES or on Product Hunt. The author advocates for a humbler, more considered approach that treats product building as an opportunity to gradually influence the future, perhaps by kicking stones in its path, rather than leaping forward for a secret advantage. This involves designing the rudder, not the boat, and recognizing that there is no inevitable endpoint in design, only hypotheses about what could come next. As Brian Eno said, the aim is to make things that belong to the future we would like to live in, not to chase hype for competitive advantage.
How ProfitBooks used Intercom to increase renewal rates by 40% - The Intercom Blog
ProfitBooks, an Indian cloud accounting startup, faced a fragmented toolchain as it rapidly grew to over 1,000 customers. They previously used separate solutions for customer support (Zendesk), email marketing (Mailchimp), feature tracking (Asana), and a self-built CRM, resulting in low customer engagement. After adopting Intercom as a unified platform for support, email marketing, and sales, their customer message volume increased 15 times, revealing previously hidden user feedback. Intercom’s automated messaging and user segmentation enabled targeted emails, boosting open rates from 7% to 45% and driving measurable business impact. By sending a payment reminder one day before account expiration, ProfitBooks achieved a 40% increase in renewal rates, and a follow-up promo converted 18% of churned trial users.
Stopping business email compromise with AI-driven email security: How VIPRE protects modern enterprises on AWS
Business email compromise (BEC) uses social engineering and impersonation rather than malicious links, bypassing traditional secure email gateways that are blind to internal traffic and lateral movement. VIPRE developed Integrated Email Security (IES), a cloud-native solution on AWS that integrates directly with Microsoft 365 via APIs, eliminating the need for MX record changes. The solution uses multiple machine learning layers, including semantic similarity models, contextual language models, and dynamic behavioral analysis of links and attachments, running on AWS services such as Amazon ECS with NVIDIA GPUs. In production, VIPRE IES processes over half a million messages daily with P95 latency between 1.4 and 78 seconds depending on sandbox analysis, detecting approximately 1% of messages as BEC attempts. Deployment can be completed in minutes rather than days, providing immediate visibility into internal risks that traditional gateways miss.
Bulk Executor for Amazon DynamoDB, Part 2: Lessons we learned the hard way
The post describes design tradeoffs and workarounds encountered while building the open-source Bulk Executor tool for Amazon DynamoDB on AWS Glue. Core requirements included real-time local-script feel, scaling to hundreds of parallel workers without user infrastructure management, and easy command extensibility. AWS Glue was chosen for serverless parallelism and a built-in DynamoDB connector, but the connector simplified data types, losing distinctions like string set versus list of strings. Challenges included estimating costs for provisioned tables, handling job termination with ^C, and streaming output via CloudWatch Live Tail which samples events and delivers out-of-order. Solutions included a two-phase cost estimate for deletes, programmatic interrupt handling, writing large output to S3 with a preview, mandating point-in-time recovery for mutating commands, and using AI for testing and linting; the team also migrated from AWS Glue 4 to 5 and 6 for faster startup and 30 percent lower DPU costs.
Top topics
- Architecture834
- Performance816
- LLMs668
- Machine Learning622
- Open Source567
- AI Agents559
- AI476
- Developer Experience448
Weekly digest
The best engineering reads, once a week.