Perplexity Launches Labs Backend Test

Today's curated insights on AI product management from X/Twitter across 60+ expert sources and YouTube channels.

Perplexity Launches Labs Backend Test

From X

AI Product Launches & Updates

Deep Research update: Arav Srinivas @AravSrinivas shared a Deep Research update on Perplexity, rolling out a new Labs backend test to 20% of users to improve performance (view: 48,252; fav: 456).
EDGAR integration improvements: Arav Srinivas @AravSrinivas announced EDGAR integration improvements in Perplexity’s Finance module, inviting users to flag issues and tag team members for fixes (view: 32,509; fav: 131).

AI Tools & Applications

Build Local AI Agents: LangChain AI @LangChainAI introduced Build Local AI Agents, enabling intelligent agents to run locally with reasoning capabilities and real-time monitoring in LangGraph Studio (view: 19,521; fav: 402).
SWE Agent: LangChain AI @LangChainAI launched SWE Agent, a LangGraph-powered system that automates software development tasks from feature planning to bug fixes with advanced codebase analysis (view: 17,114; fav: 266).
AI coding agents: Claire Vo @clairevo recommended using AI coding agents as internal communications and technical writing coaches to streamline engineering documentation (view: 11,397; fav: 124).

Product Management Insights & Strategies

PM benchmarks: Lenny Rachitsky @lennysan shared key PM insights, including Steve Jobs’ App Store intervention and Mercado Libre’s 5% PM-to-engineer ratio vs. 20–30%, offering valuable benchmarks (view: 6,071; fav: 24).
Slide deck challenges: Dharmesh @dharmesh outlined the complexity of slide deck creation, emphasizing the need for proper abstraction, engaging visuals, and clear communication when evaluating AI presentation tools (view: 6,164; fav: 71).
Interview leverage: Teresa Torres @ttorres advised candidates to assert influence in later interview stages and ensure alignment with continuous discovery practices (view: 846; fav: 4).

AI Industry Developments & News

India as AI superpower: Clement Delangue @ClementDelangue opined that India could become an AI superpower, highlighting early signs of rapid growth and innovation in the region (view: 113,630; fav: 1,502).
Mercado Libre scale: Lenny Rachitsky @lennysan highlighted Mercado Libre’s 18,000+ engineers, 30,000+ daily code deploys, and $100B valuation, underscoring massive industry benchmarks (view: 57,064; fav: 286).
Weekly AI roundup: There's An AI For That @theresanaiforit provided a weekly AI news roundup, covering ChatGPT’s recording feature, free Sora access, expressive AI audio, Reddit vs. Anthropic legal clash, Amazon humanoid robots, and Google Gemini’s automation enhancements (view: 29,682; fav: 29).

From YouTube

AI Will Disrupt This BILLION DOLLAR Industry (Google Veo): Here's How

All About AI • June 08, 2025

All About AI walks through an end-to-end pipeline using Google Gemini, Flux Pro context, Flow, Claude/ChatGPT, 11 Labs, and Suno to generate complete 10–15 second product ads—including images, video clips, voiceovers, and music—in under 30 minutes for about $10 total cost.

Key Takeaways:

Google Gemini created a base ASMR-style tech review image, which was enhanced with Flux Pro context model to add neon “Tech Girl” text, plants, warm lighting, and then upscaled with Recraft crisp.
Flow’s ingredients-to-video V2 converted the AI-generated host and Alibaba-sourced headphones into multiple product demo clips by prompting for holding, showing, and varying angles.
Voiceover scripts were drafted in Claude, spoken lines synthesized via 11 Labs, and Suno generated a chill lo-fi background track—enabling two distinct ad spots completed in ~20 minutes for under $10.

Mercado Libre: 18k engineers, 30k deploys a day, & their own fleet of planes | Sebastian Barrios

Lennys Podcast • June 08, 2025

Sebastian Barrios, senior vice president of engineering at Roblox and former head of product and engineering at Mercado Libre, details how the $100B-valued company orchestrates 18,000 engineers to ship 30,000 deploys daily and deliver 5 million packages via its own logistics fleet, driven by an engineer-led product culture and AI platform.

Key Takeaways:

Mercado Libre operates a fully vertically integrated e-commerce and fintech ecosystem with 18,000 engineers across 18 countries, 30,000 daily deployments and over 5 million package deliveries using its own planes and trucks.
Fewer than 5% of the technical team hold a traditional product manager title—engineering leaders drive both product vision and technical execution to keep teams close to user needs.
An internal AI platform called Verie abstracts data access and microservices into agentic frameworks, allowing new end-to-end features to be assembled from existing code without writing fresh code.