Opportunities › AI-powered carb counting for diabetics › Proposed app
Steady Plate
Carb counts that stay the same for the same meal, every time
- Score
- 35
- confidence 39
- Evidence
- 3
- items on the page
- Rivals
- 1
- named in sources
- First version
- 30 d
- estimate
- Type
- Mobile app
- PWA: installs on iPhone and Android
- Market since
- 2017
- Established market · AI estimate
ScreensConcept
Mobile app: People log meals on the go with photos or quick text, need consistent carb counts at meal time for diabetes management, and share data with care teams—all mobile-first everyday health tracking. Sample screens on a phone, in the proposed brand. Their numbers are sample data.
New meal
62 g
- Spaghetti, 1.5 cups48 g
- Tomato sauce, 0.5 cup9 g
- Parmesan, 1 tbsp1 g
- Garlic bread, 1 slice4 g
About this appAI proposal
Steady Plate estimates carbs for meals and gives the same number whenever you log the same food in the same portion. It is for people with diabetes who count carbs for meal planning and insulin dosing.
Instead of asking a general chat model for a fresh guess each time, it matches your meal to a fixed food entry, scales the portion, and returns a saved result with a stated range. Ask twice, get the same answer.
Your own corrections are saved and reused, so your usual breakfast stays the same number. It is a logging aid, not medical advice. Check doses with your care team.
- Tagline
- The same meal, the same carbs. Every time.
- Voice
- SteadyPlainCareful
- Why this name
- A plate you can trust to give the same answer each time. Two plain words that stress consistency, the gap in current AI carb apps.
In the first version
Members say which features they would use. It tells PWAS what people want; it changes no score.
Same food, same answer
Each food and portion maps to a stored entry. Repeat the query and the carb value does not change.
No member yet
Photo or text to structured meal
AI only reads your meal into items and portions. A fixed lookup then sets the carbs, so output is structured.
No member yet
Range shown with every count
See a low to high range and a confidence note, so uncertain meals are flagged before you dose.
No member yet
Your own food memory
Fix an item once, such as your usual bread, and it is used every time after.
No member yet
Brand identityAI proposal
Check the name against trademark registers before using it.
Palette
#0F4C4ADeep Teal · primary#FAF6EEWarm Paper · surface#1C2524Ink · text#F29A5BApricot · accent#4F9D69Leaf · data, in range
Typography
Aa
Steady Plate
ABCDEFGHIJKLMNOPQRSTUVWXYZ 0123456789
The same meal, the same carbs. Every time.
PricingAI proposal
The material gives no competitor prices. These tiers are a starting point to test, not based on market data.
Free
Free
- 10 meal lookups per month
- Saved foods
- Range with each count
Plus
$6 a month
- Unlimited lookups
- Photo logging
- Change history
- Shareable summaries
Family
$10 a month
- Up to 4 profiles
- Everything in Plus
- Caregiver share view
Would you pay Plus · $6 a month?
Your answer tells PWAS which prices people accept. It changes no score.
No answers yet.
What it could be worthEstimate
What a software business making this much sells for today, from real deals. Pick a monthly revenue.
If it makes
Worth about
$190,000
Sale range $132,000 to $263,000
$120,000 a year in revenue
From 87 software businesses listed and sold on Empire Flippers in the last year, revenue vetted by the broker: they sell for about 19× their monthly revenue, most between 13× and 26×.
Measured on profit, the middle deal is 32× monthly net profit: about 2.7 years of profit.
An estimate, not an offer or financial advice. A real price depends on profit, growth, churn, how much of the work is the owner's, and the buyer.
How it would growAI proposal
How the product would reach new customers through its own use, and how the first ones would be found. A plan to test, not a result.
- Built-in loop
- Shared outputs. Users send a weekly meal and carb summary as a link or PDF to a clinician, partner or parent. The page carries a small Steady Plate mark and a link to try the app.
- Do they become customers?
- Recipients are often clinicians or caregivers, and caregivers could become paying Family users. Test whether shared summaries lead to sign-ups and whether clinicians recommend it.
- First customers
- Adults with type 1 diabetes who count carbs and dose with multiple daily injections or a pump, reached in online patient communities.
- Where to reach them
- Diabetes online communities: Share a demo of repeat queries on mainstream chat models versus Steady Plate in forums and subreddits.
- Hacker News and tech-savvy users: Post the consistency approach; many diabetics there build their own tools and give blunt feedback.
- Diabetes educators: Offer free Family access to a few educators to try with their clients and give feedback.
- First milestone
- Fifty weekly active users logging at least 10 meals each, and ten paying Plus subscribers within 60 days of launch.
Why people already pay for this
- AI inferenceThe underlying issue is that mainstream LLMs cannot give consistent carb counts for the same food item across repeated queries
- Public claimAI carb counting apps are appearing in app stores targeted at people with diabetes
- Public claimExisting AI carb counting apps may not use mainstream LLM models because they would be too expensive for cheap or free apps
Tech stack
Every proposal builds on the same PWA base, installable from the browser on any phone or computer. Dashed boxes are what this app adds.
- AppNext.js and TypeScript
- PWAManifest and service worker
- UITailwind CSS
- Data and authSupabase Postgres
- PaymentsStripe
- EmailResend
- HostingVercel
- This app · Food dataLicensed or open nutrition databaseFixed, versioned carb values are what make answers repeatable; need a source with usable licence terms.
- This app · Meal recognitionVision/language model, structured outputUsed only to turn photos or text into items and portions, with a fixed schema and low temperature.
RisksAI inference
- RegulatoryCarb counts feed insulin dosing decisions. Depending on claims and market, the app could count as medical software. Needs legal review and clear not-medical-advice wording.
- DataA consistent number can still be wrong. Quality and licence of the nutrition database decide accuracy, especially for mixed home-cooked dishes.
- AIMeal recognition can misread items or portions from photos. Repeatability after lookup does not fix a wrong first parse.
- CommercialThe material does not name competitors or their pricing. Users may accept free apps even if less consistent, so willingness to pay is untested.
- PlatformApp stores may restrict health apps that give dosing-related guidance, and a PWA has limited camera and notification support on some devices.
Everything marked AI proposal or AI inference was written by a model from the evidence on the opportunity page. The proposed name must be checked against trademark registers before it is used.