I build AI systems + products for health & wellness brands / creators. Including myself. Miami + remote.
Miami, FL
- Tweets 7,331
- Following 858
- Followers 504
- Likes 8,457
I think AI makes building a customer app worth a closer look. My hesitation with buying: some vendors charge more as your brand's revenue grows, even when you're not getting more value from their software.
1
1
40
I tried OpenStrap to keep using my old Whoop band without renewing the subscription. A few bugs and a day’s wait later, I got a recovery score. Setup took patience, but I’m glad I can keep using the hardware.
1
16
I think jiu-jitsu gyms are relatively AI-proof. AI can't do the physical practice for you, and it can't replace walking into a gym where you know people, see friends and train together. I need that after a day of working with AI.
6
331
I'd love an avatar that reflects how I spend my time online. Hours of scrolling and rage posting would leave a visible mark. I'm curious whether seeing my “digital fitness” would give me more reason to change those habits.
1
33
I'd be reluctant to send every sensitive health question to a big AI provider. Local models can process those questions on my own computer without sending them to a cloud service. That's a reason I want local AI to get easier to use.
30
I'd like a meal and workout log I don't have to fill out myself. Photo-based macro estimates are a start. I'm imagining wearable and gym cameras logging food, sets, reps, and weights in the background. Effort would be harder to estimate.
1
2
39
I like the idea of a training app that uses WHOOP data to suggest a lighter workout after two hard days. The condition: those AI adjustments need to follow the coach's programming philosophy. Keep the expertise, and let the plan adapt to the athlete.
60
I wouldn't assume an AI profile gets better just because it remembers more. If it stores both “I love watermelon” and “I hate watermelon,” it needs a rule for which preference to keep. Memory needs maintenance, not just storage.
2
55
I asked AI to analyze my Whoop sleep history. One pattern it flagged: I stay up late, then overcompensate by going to bed much earlier the next night. I haven't tested changes yet, but that gives me a specific habit to look at beyond my nightly sleep score.
1
1
55
I’ve had clients send AI-generated app specs they haven’t read. If I just said yes, they could pay for something they don’t want. “Slop grenade” is a good name for it: generate a huge document and leave someone else to clean it up.
35
If I ran a subscription brand, I’d look at making cancellations and order checks easier for AI agents. An assistant could hunt through the customer portal, but I’m curious about giving it a direct workflow for those common requests.
38
I talk decisions through with AI and give it a specific job: help me make a committed decision so I can act. This video series came out of that process. Filming every day is one way I'm following through on my commitment to General Development.
39
I'd like to give Instacart's Clementine my order history and say: learn my tastes, suggest healthier meals, and let me commit to weekly groceries for a month. I don't know if it can do that. It's what I'd want, provided I know I'll be home for deliveries.
2
51
If I were running a supplement brand, I'd start by connecting Shopify and subscription data to AI so I could ask: “How many new subscribers did we acquire in September, and from which channels?” That's the access I'd want first.
1
25
I talked with my PT about using AI to draft exercise follow-ups from our sessions: sets, reps, frequency, and videos. He'd still review the plan. It's a workflow we're discussing, aimed at reducing his end-of-day sorting and sending.
55
If I ran a health brand, I'd stick to the major AI platforms. A niche tool may have the feature I need today, but I'd be wary of tying my workflows to it when a major could add that feature and leave me with a painful move.
1
39
