Welcome to Your Tech Moments!
AI just got a major upgrade — without any new hardware. Perplexity’s open-source TransferEngine lets trillion-parameter models run smoothly on older GPUs, slashing costs and unlocking power that used to be out of reach. This could reshape how fast teams deploy advanced AI. The gap between “cutting-edge” and “accessible” just got smaller.
Let’s brake it all down & stay informed!
News & Insights Today
Build Your Trip in Minutes — Google’s AI Mode Does It
Take Back Your Inbox — AI Tools That Actually Work
Run Trillion-Parameter AI Models on Old Hardware — Now Possible
How To: NEW Google Gemini Flows AI Agent Update
Short News in AI & Tech
AI Tools / SaaS to Checkout
AI News
Build Your Trip in Minutes — Google’s AI Mode Does It
Quick Summary
Google is rolling out new travel-features in its Search AI Mode: a “Canvas” planning workspace, global “Flight Deals” powered by AI, and agentic booking capabilities to turn plans into reservations.
Key Insights
The Canvas tool in AI Mode lets you describe a trip (dates, style, criteria) and instantly generates a draft itinerary with flights, hotels, activities and reviews.
The Flight Deals feature uses natural-language input (“I want a week of warm weather, food vibes”) and shows bargain destinations; it’s now rolling out globally to 200+ countries and 60+ languages.
Agentic booking: for U.S. users, AI Mode can now handle restaurant reservations and local appointments (via platforms like OpenTable, Resy). Flight/hotel booking is coming next, via partners including Booking.com, Marriott, Expedia.
Initially, some features (Canvas for travel) are U.S.–desktop only, opt-in via Labs. So global availability and mobile rollout will follow.
Why It’s Relevant for You
Since you work in travel and use AI tools, this update matters. With Google embedding planning + booking in Search itself, you’ll want to:
Adapt your content strategy: Travelers might use Canvas to generate itineraries—your articles and portals should show up in those suggestions.
Watch SEO implications: If AI surfaces plan drafts and deals via Google’s interface, your organic search visibility may shift.
Explore new workflow tools: You can leverage Canvas-like capabilities internally to streamline itinerary generation for your portal.
Stay ahead in travel-tech: As booking becomes agentic (AI does the booking work), the competitive landscape changes for travel services and content.
📌 More Informations: Google
Take Back Your Inbox — AI Tools That Actually Work
Quick Summary
AI-powered email tools are shifting your inbox from a never-ending chore into a strategic asset. Quartz lists nine ways to automate replies, prioritise messages, turn threads into tasks and regain control.
Key Insights
AI can draft responses for routine email types — confirmations, approvals etc — saving you hours per week.
Long email chains can be summarised in seconds, letting you skip straight to action points instead of reading everything.
Tools now learn your patterns to prioritise high-value messages, filter out noise (e.g. newsletters) and turn emails into tasks or calendar items.
The payoff: less cognitive load, fewer interruptions, more time for strategic work instead of chasing your inbox.
Why It’s Relevant for You
Since you deal with content creation, workflows and tools, this matters:
You’re likely juggling multiple email threads — adopting AI inbox tools can reclaim chunks of your time.
When writing SEO-driven articles or managing workflows, fewer distractions mean deeper focus and higher output.
Knowing these AI capabilities also gives you a content angle: you could write about how to deploy them in creative workflows or travel-business contexts.
Implementing these tools can boost your efficiency and allow you to scale your workload without increasing stress.
📌 More Informations: Quartz
Run Trillion-Parameter AI Models on Old Hardware — Now Possible
Quick Summary
Perplexity AI has open-sourced a tool called TransferEngine that enables massive language models to run efficiently across older or mixed cloud hardware, simplifying deployment and cutting costs.
Key Insights
TransferEngine enables high-speed GPU-to-GPU data transfer via RDMA, across hardware from different cloud providers (e.g., NVIDIA ConnectX and AWS EFA).
This lets trillion-parameter models (e.g., “DeepSeek V3”, “Kimi K2”) run on existing H100/H200 systems without needing the most cutting-edge hardware.
The technology is live in production at Perplexity, supporting disaggregated inference and MoE (Mixture-of-Experts) models across nodes.
It’s released open-source, allowing broader adoption and vendor-lock-in avoidance.
Why It’s Relevant for You
Since you’re working with AI tools and workflow automation (and writing about them), this is big:
It lowers the barrier to running large models, so you or your clients could experiment with advanced models without the latest GPUs.
It signals infrastructure shifts: content about AI deployment or workflow tools (which you generate) may need to discuss this kind of tech change.
For your travel-tech work: powerful models become more accessible, meaning tools you write about (agentic booking, itinerary generation) could adopt them sooner.
Efficiency boost: If you ever build or deploy an AI tool, knowing that hardware demands are easing helps plan budget/workflow.
📌 More Informations: Medium Datadriveninvestor
How To
NEW Google Gemini Flows AI Agent Update
Quick Summary
This video shows how Google Gemini’s new Workspace Flows let you build AI agents that automate email, summaries, reminders, and content tasks directly inside Google Workspace. No APIs, no Zapier-style wiring — just templates, schedules, and optional custom Gems that run your work in the background.
Key Insights
Workspace Flows is a new Gemini feature that lets you build AI-powered workflows (agents) across Gmail, Chat, Calendar, Docs and third-party apps.
You can start from ready-made templates to summarize emails, notify you about key senders, create pre-meeting briefs, or send daily unread-email digests.
Flows support conditions, schedules, and multi-step chains: filter emails → extract action items → summarize themes → send a Chat notification or email.
You can plug in custom “Gems” (your own AI agents) to generate content, such as daily SEO blog posts or fully drafted email replies.
What Can I Learn?
How to turn on Workspace Flows inside your Google Workspace and access the templates.
How to build a flow that watches your inbox, summarizes key emails, and pings you in Google Chat.
How to schedule recurring workflows (e.g., every morning at 8 a.m. summarize yesterday’s unread emails).
How to connect a custom Gemini Gem (like an SEO content writer) into a flow and feed it variables like unread emails.
Which Benefits Do I Get?
Less manual email triage: AI filters, labels, and summarizes your inbox for you.
Fewer tools and subscriptions: many “Zapier-style” automations can live inside Workspace at no extra cost if you’re on a Gemini plan.
Faster content and operations: auto-generate SEO posts, reminders, and team updates without copying data between tools.
Clear visibility: built-in activity logs show what ran, what failed, and history for around 40 days so you can debug and improve.
Why It Matters
Workspace Flows turns Google Workspace from “place where your work lives” into “system that quietly does work for you.” Instead of wiring multiple tools with APIs, you can build agentic workflows fully inside Google’s ecosystem. For teams spending hundreds per month on Zapier-style automations, this offers a cheaper, more integrated alternative that scales with your existing email, docs, and calendar usage. It also lowers the barrier to real AI agents: non-technical users can drag-and-drop powerful flows instead of writing code.
If you live in Gmail and Google Docs, this is a direct upgrade to your day.
You can:
Turn chaotic inboxes into structured summaries and action lists.
Automate daily SEO or content tasks directly from templates + custom Gems.
Replace parts of your Zapier stack with native, cheaper Workspace automations.
Build shareable workflows for your team so everyone benefits from the same AI agents, not just power users.
📌 Watch the full Video: YouTube
Short News in AI & Tech
1) HBR – AI Tools Transform Market Research
AI is reshaping market research with synthetic personas, faster insight generation, and cheaper testing cycles. Companies use AI to run simulations, refine messaging, and understand shifting consumer behavior without long surveys. The industry moves toward “decision-makers augmented by AI,” speeding strategy and reducing guesswork.
📌 Read More: HBR
2) Databricks – PDF Parsing for Agentic AI
Databricks says PDF parsing is still a major bottleneck for agentic AI. Their new ai_parse_document tool replaces fragile multi-step pipelines with a unified system capable of understanding layout, structure, tables, and visuals. This gives autonomous agents high-fidelity access to documents, boosting reliability across enterprise workflows.
📌 Read More: Venturebeat
3) beehiiv – Digital Products & AI Tools for Creators
beehiiv evolves into a full creator platform, adding digital product sales, AI website tools, podcasts, analytics, and more. Creators can sell ebooks, courses, templates or memberships without paying a commission. The goal: become the operating system for modern content businesses and expand far beyond newsletters.
📌 Read More: Netinfluencer
4) InfoWorld – JavaScript Tools for AI Development
JavaScript now offers a broad toolkit for AI, from TensorFlow.js to Transformers.js and Brain.js. Developers can train or run models directly in the browser or Node environment. This unlocks machine-learning capability for web developers without switching to Python, helping teams embed AI directly into existing web apps.
📌 Read More: Infoworld
AI Tools / SaaS to Checkout
1) Tenki
What it does: AI-powered forecasts for prediction markets (e.g., Polymarket, Kalshi). Upload data/markets, get probabilistic predictions.
Why it’s useful: For traders or analysts—it gives “what might happen next” signals instead of just historical data.
👉 Try it here: Tenki
2) Plugin.st
What it does: Describes itself as “Build Plugins, Add-ons & Desktop Apps. Describe what you want to build, and get ready-to-use code in seconds.”
Why it’s useful: If you’re creating custom tools (desktop apps, browser extensions, WordPress plugins) this lets you skip a lot of boilerplate.
👉 Try it here: Plugin
3) Nebius Token Factory
What it does: A production-grade inference service for open-source models. Sub-second latency, autoscaling, predictable cost, no-infrastructure-wrangling.
Why it’s useful: If you build AI tools or services and need scalable model inference (text, code, vision) this offers a ready platform.
👉 Try it here: Nebius
4) DatePhotos.AI
What it does: Upload 5-20 selfies, and it generates 80-180 AI-optimized dating profile photos (for Tinder, Bumble, Hinge). Features include “Realness Score”, 20-30 minute delivery.
Why it’s useful: If you’re in the content/creator space this shows how AI can serve niche photo-services; also useful example of monetisable AI micro-apps.
👉 Try it here: DatePhotos
🙏 Thanks for reading till the end — you’re the reason this newsletter exists.
Got a thought or idea? Hit reply — I read every message.
⭐⭐⭐⭐⭐ Your feedback helps make each issue sharper.
Not subscribed yet? Join here & share it with someone who’d enjoy it too.
See you in our next edition!
Adrian