Nathan Weill here.
I'm the CEO @ Flow Digital, where we help companies unleash their full potential by strategically automating every inch of their workflows. Each week, I share hot AI and automation tips to help you move your business into the future successfully.
Pen.dev Puts a Design Team on Your Canvas
Our team has been using it as an on-call UI specialist, and the design-to-code handoff looks different because of it.
Design is the bottleneck between "we need a portal" and "the portal is live." Interfaces, content cards, and app refactors sit in a queue waiting on a designer, then wait again for someone to turn the mockup into working code.
Pen.dev is an agentic canvas that targets exactly that gap. Connected to your Claude or ChatGPT account, its AI agents design directly on the canvas and can carry the work through to code.
An agent designs an app interface on Pen.dev's canvas, with the model selected in the prompt bar.
Here is what one of our team has found using it:
A UI specialist on call: they've run it like their own designer for interfaces, content cards, and portals, and used it to refactor an existing app with strong results.
It opens the pull requests itself: connected to their terminal, the tool handles the pull requests, leaving a person to review and merge.
Reusable brand agents: you can save a trained specialist (say, a "Flow Digital UI Designer") and select it so new screens come out in the styles and components you've chosen.
For an operations leader, this shifts who can produce a usable interface, and keeps a human on the final approval.
User feedback (from our own use)
What's working: refactoring an existing app produced strong output; it runs as a hands-off UI specialist and opens its own pull requests; saved brand agents keep screens on-style.
Still early: our use is recent, a person has to review and merge every pull request.
See what Pen.dev's agentic canvas can do.
Gong's New Enrich Feature
Stop losing deals to missing phone numbers and dead email addresses.
Gong Enrich fills a missing contact in one click, with Agent Builder triggers and a data-partner waterfall behind it.
Reps burn hours every week hunting down a prospect's correct email or direct line, or they skip the outreach when the record is too thin. Gong Enrich closes those gaps inside Gong, so your team moves straight from research to conversation.
The feature pulls verified contact and account data into Gong the moment a record is incomplete, then keeps it current as the account changes.
What your team can do with it:
Fill a missing email or phone number automatically before an outreach sequence stalls.
Enrich a contact and its parent account in one click, without jumping between tools.
Keep records fresh on their own: refresh data when the CRM changes, a date passes, or an AI Tracker fires.
What that gets you:
Fewer bounced emails and wrong numbers dragging down sender reputation.
Reps selling instead of data-cleaning, with context ready before the first touch.
A CRM that stays accurate without a manual scrub, since enriched data syncs back.
Freedom to use the data contracts you already pay for (Apollo, ZoomInfo, Lusha, and more) or buy on demand with Gong Credits.
The payoff is cleaner data and more of your reps' time spent actually talking to buyers.
Beat AI Sameness in Your Content
Point automation at the inputs only you have, so your content stops reading like everyone else's.
Every team now runs the same tools on the same prompts, so content all starts to read the same: same structure, same three tips, same sign-off. When your blog sounds like your competitor's, buyers skim past it.
SurveyMonkey saw the upside of breaking the pattern, with a 73% jump in click-through when it traded generic "surveys are powerful" ads for messaging built on what one audience actually cared about.
The fix is simpler than writing more: feed your content engine raw material competitors cannot copy. Automation is the fastest way to surface it.
Start with the inputs closest to your customers:
Mine your own calls: run your Fathom or Gong transcripts through AI to pull the exact pain points, objections, and words your customers use, then turn each into a post.
Map the open ground: track competitor blogs, sitemaps, and changelogs automatically, and get alerted to the topics no one in your space is covering.
Catch themes early: scrape the LinkedIn hashtags your buyers follow so you publish on a rising topic before it gets crowded.
Pipe all three into a Zapier flow and the material lands in one place each week, with no manual digging.
Your calls are the start. The richer sources are the ones sitting in your own systems and out in public, waiting to be read at scale.
Publish research only you can run: you already sit on data no competitor has, like your booking patterns, survey answers, and product usage. Automate the pull and analysis, and one dataset becomes a first-mover trend report, a run of posts, and a webinar. This is exactly how SurveyMonkey built its State of Curiosity report.
Mine the reviews everyone ignores: your buyers leave candid, unprompted feedback on G2, Capterra, Reddit, and niche forums. Automate the review pull and theme-tagging to surface unmet needs and the real language customers use, often the earliest signal of a topic before it trends.
Turn lost deals into content: feed your win/loss notes and CRM comments through AI to spot the objections and comparisons buyers actually weigh, then answer them head-on in public posts and comparison pages.
Keep one human in the loop. Automation surfaces the raw material fast, but a person picks the angle, adds the point of view, and decides what you are willing to say that a competitor will not. That judgment is what keeps the output sounding like you instead of a model's average.
The teams pulling ahead are not writing more. They are writing from inputs no one else holds, and letting automation do the digging.
A Filter for the Endless Flood of New Tools
A repeatable filter for the ~44 new AI tools that launch every day, so you adopt the one that pays back and skip the 43 that don't.
Around 44 new AI tools go live every day, and directories now list more than 50,000 of them. The risk isn't missing one. It's that chasing them quietly drains budget and focus: the average company already runs 305 SaaS apps and leaves 53% of its licenses idle, which runs to roughly $865,000 a year in unused software at a 100-to-500-person company.
So the skill worth building is a filter, not discovery. More tools rarely means more output. The teams that stay current without the waste treat tool-watching as a light, scheduled habit, then hold every new option to the same test before it touches the stack.
Where to watch, in 15 minutes a week:
Pick two signal sources, not ten: a daily roundup like The Rundown AI or TLDR AI for what's moving, plus Product Hunt's AI category for what's launching.
Keep a living shortlist, not 40 open tabs: log anything worth a look in a shared base like Airtable so the whole team feeds one list.
Let a real problem pull the tool: start from a bottleneck you already have, then go looking, never the reverse.
The five-minute vet before you commit:
Tie it to one job and a number: name the task it replaces and the hours or errors it should save, then run the free trial against that, not the polished demo.
Check how it connects and how you leave: confirm it fits your current stack (does it talk to your CRM, your sheets, your Zapier flows?) and that you can export your data if you drop it.
Weigh the real cost, not the sticker: add seats, onboarding time, and the admin to maintain it, because more than half of SaaS licenses end up unused, usually from tools bought on impulse.
Confirm it's safe to trust: look for basic security proof (SOC 2, clear data handling) before anything sensitive flows through it.
Where a human still decides: software can surface and even trial a tool, but a person should own the go/no-go, the security sign-off, and the quarterly cull of what's stopped earning its seat. Automate the watching, keep the buying deliberate.
We put this to work for clients, auditing noisy stacks and connecting what's left into one system that runs itself (see the case studies).
Want us to do it for yours? Book a 20-minute discovery call →
SMART WORDS OF THE WEEK:
— William James
With 40+ tools launching a day and every team running the same prompts, the edge is no longer what you can add, it's what you have the discipline to ignore.
 |
Stay in touch with our team from anywhere.
