Claude and Notion AI agents, IKEA's €1.3B with AI

A product manager describes how he uses Claude-driven agents, like podcast, sales, and coding managers, to auto‑generate prep tasks, post‑meeting follow‑ups, and weekly retrospectives, letting him wake up to a to‑do list already built. The system leverages the OpenClaw harness while keeping costs and access safe.

Ryan Nystrom explains Notion's AI-driven spec‑first workflow, where engineers dictate ideas, Codex writes specs, and custom agents generate pull requests, cutting CI time and automating standup prep. The system lets engineering managers stay hands‑on while agents handle routine coding, accelerating delivery and reducing burnout.

Users quickly spot static login screens, treating prototypes as fake and skewing test data. This Smashing Magazine guide shows how to build a realistic, interactive login flow in ProtoPie, complete with text inputs, validation, error handling, and a Face ID animation, without writing code, ensuring more authentic usability feedback.

The Smashing Magazine article presents ten research‑backed findings that quantify how user experience directly affects business outcomes, from a 1:100 cost saving when fixing design early to a 20% conversion drop for each second of load delay. It shows that measurable UX improvements translate into higher revenue, retention, and growth.

The article shows why spinners and generic “loading” messages mislead users when AI agents pause to think, and introduces clear status‑update microcopy and visual containers that explain the AI’s internal steps. These patterns turn wait time into transparent, trust‑building moments.

While tech firms cite AI to slash staff, IKEA used its AI chatbot to surface design‑focused customer queries and retrained 8,500 call‑center agents as remote interior design consultants, generating €1.3 billion in new revenue. The post shows how AI can reveal hidden value work instead of just cutting jobs.

Aikido rebuilt the cybersecurity buying experience around a product‑first, self‑serve model, publishing transparent pricing and letting users try and buy without sales calls. By cutting friction and alert‑noise, the startup grew to over $40 million ARR in just 3½ years, proving PLG can succeed in a traditionally sales‑heavy market.

Intercom explains how AI sales agents can respond to prospects instantly, removing the historic speed‑to‑lead bottleneck. With 24/7, multilingual engagement, sales organizations can shift SDRs from triage to relationship‑building and overhaul metrics that once focused on response time.

Dan Shipper predicts that AI‑driven agents inside tools like Claude Code will become the central hub of work, boosting the value of product managers and full‑stack designers while rendering traditional CLIs obsolete. He also argues SaaS remains strong and the forward‑deployed engineer will be the most critical hire.

Felix Rieseberg, Anthropic’s Claude Cowork lead, demonstrates three practical Claude workflows: converting floor plans into interactive 3‑D walkthroughs, automatically logging his Twitter promises, and building a cheap $20 hardware “buddy” that lets him approve Claude actions with a button. These examples showcase Claude’s potential for automating everyday tasks.

Anthropic engineer Thariq Shihipar explains why HTML has replaced Markdown for prompting Claude Code, enabling richer visual planning, throwaway micro‑UIs, and living design systems that travel with the codebase. The approach lets AI agents allocate compute to planning and interfaces rather than raw production code.