Founder & CEO, OffDeck: the modern partner for building enduring companies. Strategy, growth, RevOps and engineering for tech and AI teams.
New York & Cincinnati
Joined June 2009
- Tweets16.7K
- Following1.4K
- Followers3.5K
- Likes54.2K
Pinned Tweet
I live in two worlds: shipping code and fixing go-to-market.
That's OffDeck. We work alongside tech and AI teams on the CRM, data and systems that growth runs on, so founders can get back to the product.
If that's your bottleneck, my DMs are open.
offdeck.com
The most expensive line in a services contract is "ongoing support as needed."
Nobody knows what it means, so nobody knows when it's done.
OpenAI's 12-hour prompt cache costs 2.5x to write.
The 30-minute default is 1.25x. Reads are 0.1x either way.
So it only pays if you'd rewrite that prefix 3+ times a day. At two it's a wash. A loop that runs constantly never pays the rewrite.
Who's actually in that band?
Changelog: developers.openai.com/api/do…
prompt_cache_options.ttl = "12h", GPT-5.6 and later, Responses API. Default stays "30m". You can also set ttl per prompt_cache_breakpoint, so a stable system prefix gets 12h and volatile context stays on 30m.
Reads are 0.1x on most models, 0.05x on GPT-6.1 Sol. Sweet spot is bursty traffic with a big stable prefix and long idle gaps — not a hot loop, not a cron job that fires twice.
Brandon Lipman retweeted
Want to share my thoughts as one of the few, if only people, that have deployed AI Agents that are moving money and making payments within a banks firewall for their customers.
Banks are shutting off access to external AI Agents (for now). Why did this become such a problem all of a sudden?
This has been happening since the beginning of the year. When OpenClaw came out it was maybe a few hundred hackers/engineers that set their OpenClaws up and gave it access to their bank accounts. Why?
Not because they're crazy, but because they hosted the agent, they trusted it, and they saw the power for how much these agents could help them do both mundane and complex tasks, which banking has plenty of.
Fast-forward to the last two months with Instinct, Muse, and the 100+ personal AI Agents that have come out and that number has gone from a few hundred, to thousands.
Banks were not ready for this activity, and quite frankly did not expect this level of agentic activity to come this fast. So they did the right thing, shut the gates.
I don't think this will be the case forever, and banks will need to support this in a safe and trusted way where most importantly THE BANKS control the experience.
Here are my thoughts on how this all plays out, if you think a banker or anyone at any Financial Institution would benefit, please share. Thank you!
paymanai.com/blog/ai-agents-…
I’m getting married in 14 days. I feel like a very lucky man. Can’t ask for a better woman to spend the rest of my life with.
Brandon Lipman retweeted
Bragging about how much software you’re shipping with AI is like holding down the shutter button and bragging about how many photos you took.
Brandon Lipman retweeted
Talking to smarter folks than me, I'm convinced many of the AI folks in my timeline are full of shit.
Nobody is "running 20 agents over night" and building stuff for actual users. Maybe some are building internal tools or disposable software. Maybe.
But building software people like using? That doesn't get hacked on day one or blow up after the 3rd user? Nope.
I don't even understand what that's supposed to look like. Do you work out a 57 pages document that perfectly describes what you want to build and then summon 14 agents and have them run wild for 6 hours? And what comes out on the other end isn't a broken pile of shit?
Nope. Not buying it.
PS: it may also be that I have an IQ of 82 and can't figure it out.
If you have ever used Cursor IDE, please go check your bank statements. A huge volume of fraudulent charges to a card that was never connected to Cursor.
If you have ever used Cursor IDE, please go check your bank statements. A huge volume of fraudulent charges to a card that was never connected to Cursor.
Brandon Lipman retweeted
Today we're launching Glaze 💠
Create any desktop app in minutes by chatting with AI.
Beautiful, powerful, and truly personal.
Learn more on glazeapp.com
Follow @glazeapp for updates.
Brandon Lipman retweeted
ARR is based on annualized contracted recurring revenue, not simply multiplying one random month’s revenue by 12.
I created a GitHub repo for wedding planning. It's great to have a soon-to-be wife who is technical. 💘
The net creation of productivity is now basically becoming a commodity within technology due to AI.
Most of the world is still not actually putting AI in prod, even though everyone says they are. Whats being said publically vs whats real are very different realities.
Hallucinations are going down sure. Our ability to use AI to drive value is in our ability to think and ask the questions which will enable it to solve actual problems.