When I ask for an introduction, I’m asking someone to put a little of their reputation behind it.
It matters that they understand why both people would want to talk.
What’s an introduction that made a difference for you, and what made it work?
15
A former colleague launches a company.
You remember talking it through when they weren’t sure about leaving their job.
You haven’t spoken in months, but you’re glad they went for it.
“Really glad to see this. How’s it going?”
There’s a whole conversation behind that message.
13
Some relationships don't fade because we stop caring.
They fade because our routines change.
When someone moves or changes jobs, we lose track of their lives, not just touch.
Maybe staying connected isn't about finding more time, but staying aware of the people who matter.
1
2
83
An introduction takes five minutes. What follows can shape the next five years.
A future cofounder. A first customer. Someone who helps you see a problem differently.
It’s easy to remember the conversation and forget the person who made it possible.
22
I’ll catch up with someone I care about and think, why did we wait so long?
Nothing happened. Life just got busy.
I’d love for AI to nudge me: “Remember that idea they couldn’t stop talking about? Ask if they’re building it.”
I’d find that useful.
1
1
31
Meruzhan Danielyan retweeted
Today we’re launching Sage - your 24/7 AI recruiter.
Sage works just like a human recruiter, except she works around the clock - even while you're sleeping. Sage hand-picks the best opportunities for you the moment they come out, prepares your application packet so you can apply with one tap, and runs mock interviews and gives you feedback.
The best part: you interact with Sage simply by text.
Try Sage now → saywise.com
Software failures aren’t new.
What’s nee is how hard they are to explain once AI gets involved.
Bugs didn’t disappear, the causal chain did.
#SDD #SpecDrivenDevelopment #SoftwareEngineering #AI
29
Why do big companies love AI for prototyping
but ban it in production?
Because AI builds demos fast —
not systems with context.
When prod breaks, engineers debug code they didn’t author.
That’s the real problem.
#SDD #SpecDrivenDevelopment #SoftwareEngineering #AI #Context
32
Software isn’t slow anymore.
Confidence is.
AI killed the cost of writing code.
It didn’t kill the cost of knowing if it works.
That gap is where velocity breaks.
#SDD #SpecDrivenDevelopment #SoftwareEngineering #AI #FutureOfWork
27
Spec-driven development is real.
Not “coming soon.” Not “experimental.”
It’s already here and our tooling hasn’t caught up. We’re writing intent, not code.
But we’re still verifying like it’s 2016.
That gap is where confidence breaks.
#SpecDrivenDevelopment #SoftwareEngineering
26
Debugging AI-generated behavior with logs feels like:
Trying to reconstruct a car accident
using only speedometer readings. #SoftwareEngineering #FutureOfWork #SDD #AIEra
35
Shipping AI code without verification is gambling. #SoftwareDevelopment #specdrivendevelopment #sdd #aiverify #verification
26
Managers don’t want more dashboards.
They want defensible answers:
- What broke?
- Who was affected?
- How bad is it?
#SoftwareEngineering
9
AI didn’t just change how we write software.
It broke the feedback loop that used to verify it.
We now debug behavior we didn’t author, because code is generated, not built.
That’s not a tooling gap.
It’s a verification gap.
Every abstraction shift added safety.
AI will too.
9
Nothing like airline “partnerships.”
@united sells you a @lufthansa flight.
Lufthansa flies you to Frankfurt.
You try to pay for their lounge.
Lufthansa: “Nope, United sold the ticket.”
@staralliance : Together, but not that together. 🤷♂️
1
52
Meruzhan Danielyan retweeted
Starlink now active in Armenia!
Starlink is now live in Armenia, marking the 130th country, territory or market around the world where Starlink’s high-speed internet service is available 🛰️🇦🇲❤️ → starlink.com/map?country=AM
Meruzhan Danielyan retweeted
🚀 Deep Lake 4.0 is here!
We're super excited to announce:
🔥 2x cost-efficient queries with our new index-on-the-lake
🔍 Supports Lexical Search (BM25), Semantic Search (ANN) and Late Interactions (MaxSim in alpha)
🔄 Eventual consistency for seamless data handling.
🧩 No more dependencies aside NumPy
🛠️ A simpler API for smoother development including async support ... and so much more!
Experience the future of data lakes today! #DeepLake #DataScience #BigData
Meruzhan Danielyan retweeted
We’re giving away 5 free tickets to our sold-out Founder's Night sponsored by @WeWork and @helpbnk this Tuesday at #LATechWeek!
To enter, just follow us + @TechWeek_ and share this post!
Founders Night Event 👉 lu.ma/gei8i8rl
