AI project · Consumer health · Working MVP

Ridge Health

One small win today is enough.It's the first thing you see when you open the app, and honestly, the whole idea.

A voice-first nutrition and fitness coach for people who have quit tracking apps before. You speak your meal, get a plan built around the food you actually cook, and when life gets in the way, the plan adapts instead of making you feel like you failed.

my favourite part is the evals section ↓

RoleFounder, PM, designer, builder
MarketsIndia and US
AIVoice, meal ideas, auto-adjust, evals
StatusWorking MVP, in iteration
Weekly diet plan with Indian meals and macros
Home screen with Voice Log Meal and Today's Focus
AI Meal Ideas with mode and cuisine choices
ProblemThe quit point

Tracking apps lose people once logging feels like homework and one bad day feels like failure.

InsightEmpathy is the gap

No major app can tell a lazy day from a sick day, a trip or a festival.

BuiltA coach that adapts

Voice logging, one-tap meal confirm, rest and treat days, pause, and weekly auto-adjust.

MeasureStill logging in week 4

North star: users with 5+ logged days a week, a month in.

01The problem

Why people give up on food tracking

Every tracking app has the same drop-off curve. I call it the disengagement gap: the moment the novelty wears off and logging starts to feel like homework. For people who eat home-cooked Indian food, it comes even sooner, because the apps were never built for a thali.

Week 1Motivated

Logs every meal, weighs the rice, checks macros twice a day.

Week 2Friction

Searching for "aloo gobi" returns five entries with five calorie counts.

Manual entry fatigue
Week 3Life happens

A work trip, a cold, a wedding. The plan doesn't know.

Rigid plan
Week 3Guilt

The streak breaks. The app shows a red number and nothing else.

No empathy
Week 4Ghosted

The app stops being opened. It never asks why.

02Who it's for

Who it's for: people starting over (again)

P
Priya, 34
Primary · busy professional
Context
Tech job in the Bay Area. Cooks Indian food at home, travels twice a month.
History
Tried MyFitnessPal and HealthifyMe. Quit each within a month.
By the time I find the right dal entry, I've lost the will to log dinner.
A
Amit, 41
Secondary · health scare
Context
Lives in Pune. His doctor told him to lose 8 kg. Eats what his family cooks.
Needs
Guidance that fits a family kitchen, not a bodybuilder's meal prep.
I can't weigh my food. Just tell me what to eat tomorrow.
M
Meera, 29
Edge case · irregular weeks
Context
Consultant with weeks on the road and weeks at home. Gets sick a few times a year.
Needs
A plan she can pause without starting over.
Every time I travel, I lose my streak and my motivation.

Personas are illustrative.

03Competitive research

How MyFitnessPal, HealthifyMe and Fittr compare

I analysed the three most-used tracking apps among my target users on two things: how logging feels, and how each one keeps people accountable. Analysis as of late 2025

AppHow you logEffortSpeedAccuracyAccountability model
MyFitnessPalThe databaseSearch and barcode scan. Great for packaged food; you guess portions for everything else.ModerateFastMedium
Crowdsourced entries
Self-accountability. Reminders and streaks. If you stop, it doesn't notice.
HealthifyMeThe AI snapPhoto, text and voice. Handles multi-dish thalis well.LowestFastestMedium
Misses hidden ghee and oil
AI nudges. A coach that pings you, plus glucose monitors on the premium tier. Easy to snooze.
FittrMacro precisionManual search of raw ingredients, weighed in grams.HighestSlowestHighest
Verified entries
Human coaches and peers. Hardest to ignore, but only as good as your coach, and intimidating for beginners.
Ridge HealthThe adaptive coachVoice first, plus one-tap confirm of meals already in your plan.LowFastMedium
Plan-anchored portions
Empathetic adaptation. Rest days, treat days, pause, and targets that adjust to how your week actually went.

Where every app falls short

MyFitnessPal · HealthifyMeThe empathy gap

Can't tell a lazy day from a mental-health day, a fever or a trip.

Ridge: Rest and treat day calendar, Pause My Plan, "do what you can, we'll adapt."
MyFitnessPal weakestThe integrated-life gap

Plans ignore what actually happened last week.

Ridge: Auto Adjust recalculates next week's targets from last week's actual intake.
All threeThe maintenance gap

Accountability ends at goal weight, so the weight comes back.

Ridge: Habit-first framing ("consistency wins") Maintenance mode is on the roadmap..
All threeThe verification gap

You can log a salad while eating a pizza.

Ridge: Not solved. I chose not to fake it. Plan-anchored logging reduces the guesswork instead.

Where Ridge sits

04SWOT

Ridge's SWOT

SStrengths

  • Lowest-effort logging: voice, plus one-tap confirm of meals already in the plan.
  • Built for Indian kitchens: daliya, roti with dal and sabzi, tofu with brown rice, with India and US regions.
  • Empathy as a feature: rest days, treat periods and Pause My Plan are core, not add-ons.
  • Adaptive targets that respond to how the last week actually went.

WWeaknesses

  • AI reliability: in testing, meal ideas failed when both AI providers were down at once.
  • Accuracy ceiling: voice estimates miss hidden oil and ghee, the same issue photo logging has.
  • No human layer yet: Coach and Community are still "coming soon."
  • Solo-built MVP with no proven retention data yet.

OOpportunities

  • The empathy gap is wide open. No major player adapts to sick days, travel or festivals.
  • Indian diaspora in the US: Western databases serve them badly and Indian apps are built for India.
  • Hybrid AI plus human coaching as a premium tier, combining Fittr's stickiness with AI's scale.
  • Maintenance mode for life after the goal, which every competitor drops.

TThreats

  • HealthifyMe already has voice and photo AI and can copy empathy features.
  • MyFitnessPal's database is a moat for packaged food.
  • Notification fatigue: AI nudges get snoozed, and Ridge isn't immune.
  • LLM cost per user and health-advice regulation as the product scales.
The insight
I'd rather people log roughly every day than perfectly for one week.
The product bet

Optimise for consistency over precision. Make logging nearly effortless, and make a bad day cost nothing. If people keep showing up, accuracy can improve over time. If they quit, nothing else matters.

05The product

Walking through the app

Ridge Health home screen 1 2 3 4 5
1
Lower the bar on purpose

"One small win today is enough." The greeting sets expectations before the user sees a single number.

2
Voice is the front door

"Voice Log Meal" is the biggest button on the screen. Say "two rotis and a bowl of dal" and it's logged.

3
Room to grow

Coach and Community are visible but greyed out. They signal the roadmap without crowding the MVP.

4
Today, not forever

"Today's Focus" asks for three meals and one workout, with the promise "Do what you can, we'll adapt."

5
Proof from people like you

"Consistency Wins" celebrates streaks and small milestones, not dramatic before-and-afters that intimidate beginners.

6
Plans in your kitchen's language

Vegetable daliya for breakfast. Three rotis with dal and sabzi for lunch. Portions in bowls and rotis, not grams, with macros shown underneath.

7
Quick Confirm

If you ate what the plan said, tick Breakfast, Lunch and Dinner, or hit "Log All Meals." Logging a planned day takes about three seconds.

8
"Need variety?"

Every meal has a one-tap swap, so boredom doesn't become a reason to quit.

9
Auto Adjust, one switch

Turn it on and next week's targets learn from this week. More on that below.

Weekly diet plan with quick confirm and Indian meals 6 7 9
Account screen with rest and treat days calendar and Pause My Plan

Designed for real life, not perfect weeks

10
Rest and treat days calendar

Mark a Diwali weekend, a birthday or a recovery day in advance. They count as planned days off, so the streak survives.

11
Pause My Plan

Sick, travelling or simply overwhelmed? Pause instead of quitting. Coming back is one tap, with no restart and no lost history.

12
Personalisation that unlocks value

Goal, dietary preferences, meal frequency and activities shape the plan. The prompt explains what you get for filling them in.

Progress that encourages instead of judging

13
"What You're Good At"

A strengths-based snapshot that unlocks after a few days of logging. It leads with what's working, not what's missing.

14
Empty states that guide

Day one shows "Ready to start" and "First week loading," never a wall of zeros labelled as failure.

15
AI engagement as a metric

Voice conversations are counted and shown, which is how I track whether the voice-first bet is working.

Progress screen with strengths snapshot and recommendations
Onboarding with India and US region choice

Onboarding in under a minute

16
Region first

India or US decides the food library, units and meal suggestions from the very first screen.

17
Only what's needed

Age, gender, height and weight are enough to set a starting calorie target. Everything else can wait until the user sees value.

06Where AI does the work

The three AI features

Auto Adjust: the plan learns from your week

Plan adjustment details: last week target 12,600 kcal, actual 3,600, plan adjusted for this week
Compares plan vs. reality

Last week's target (12,600 kcal) against what was actually logged (3,600 kcal).

Explains itself

"Last week was 71% under target" sits right above the change, so the user knows why.

Shows every day

Original, adjusted and change, day by day. No black-box changes to someone's diet.

What I'd change next: a 71% shortfall is more likely missing logs than under-eating. Before raising targets, Auto Adjust should check logging completeness and ask "Did you eat more than you logged?" Raising calories because someone forgot to log would be the wrong call.
AI Meal Ideas with Variety and Quick and Easy modes and cuisine picker

AI Meal Ideas: never stuck on "what do I eat?"

  • Two modes match real moods: Variety for new flavours, Quick & Easy for minimal prep after a long day.
  • Nine cuisines, Indian first in the list, so suggestions fit what's in the fridge.
  • Allergies and avoid list sent with every request, so the model never suggests peanuts to someone allergic.
  • Three ideas, not thirty. Fewer choices means a faster decision.

Voice logging and Ask Anything

  • Natural speech becomes structured meals with portions and macros.
  • "Ask Anything" is a chat coach for questions like "Is poha okay before a workout?", with guardrails that keep it to wellness, not medical advice.

Built for failure

Requests go to Gemini 1.5 Flash first, with GPT-4o-mini as the fallback. Testing surfaced a case where both failed at once. The fix on my list is a third layer: a curated library of pre-written meal ideas, so the user always gets an answer.

07Voice AI

Why I went voice-first, and what made it hard

Why I built voice first

  • Effort is why people quit. My competitive research showed manual entry fatigue as the top reason people drop MyFitnessPal and Fittr. Speaking a meal takes seconds.
  • Indian meals don't fit search boxes. "Do roti, ek katori dal aur thodi sabzi" is one sentence out loud, and six searches in a database.
  • Hands are busy. People log while cooking, eating or walking. One tap to speak beats typing on a phone.
  • One input for everything. The same "Tap to Speak" button logs a meal, water, steps or sleep, or asks a question.

How it works

Tap to Speak→ Speech-to-text→ Intent: log or ask?→ LLM extracts items + portions→ Nutrition lookup→ User confirms ✓

Questions go to Chat with Ridge AI, which has specialised flows for the moments people usually give up: rigid diet plans, cravings, losing motivation and holiday planning. Allergies and busy-schedule flows are next.

Today's Logs with daily snapshot and a Tap to Speak button
Ridge AI chat options with specialised flows for cravings, motivation and holidays

Challenges I hit

ChallengeExample
Mixed-language speech"Ek plate poha with chai"
Vague portions"Thoda sa rice," "dedh katori dal"
Several things in one sentence"Had lunch, drank 2 glasses of water, walked 20 minutes"
Being misunderstood"Didn't understand" was the top thumbs-down reason in my feedback data
LatencyVoice feels broken if the reply takes more than about 3 seconds

"Didn't understand" being the top thumbs-down reason is why a Hinglish evaluation set is first on my list.

08AI evaluation pipeline

How I evaluate the AI

Ridge has a built-in admin console that tracks every AI feature on three layers: what users think of it, how the models behave, and whether the service meets its targets. With only a handful of beta users, it already caught the most important problem in the product.

33LLM calls traced
2providers compared
4SLOs monitored
27%feedback response rate
7console views
Layer 1 · UsersIs it useful?

Thumbs up or down on every AI experience, a reason when it's down, and results broken out by feature.

  • 22 completed AI experiences
  • 6 ratings, a 27.3% response rate
  • Voice logging: 8 uses, 6 ratings, 17% positive
  • Top complaint: "Didn't understand"
Layer 2 · ModelsIs it working?

Every call is logged with feature, provider, latency and errors, so the two models can be compared side by side.

  • Gemini 1.5 Flash: 18 calls, ~327 ms
  • GPT-4o-mini fallback: 15 calls, ~2,460 ms
  • Daily call volume and recent errors
  • Prompt versions tracked separately
Layer 3 · SLOsIs it good enough to ship?

Four service-level objectives with explicit targets, checked against live data.

  • p95 latency under 3,000 ms
  • Error rate under 5%
  • Tool failure rate under 2%
  • Cost per task under $0.05
Admin overview: events, active users, LLM calls, latency and feedback
Overview
Feedback tab: thumbs up rate, reasons and per-feature breakdown
Feedback by feature
LLM Ops: calls per provider, latency, daily volume and recent errors
LLM Ops
SLO tab: 2 of 4 targets met, with p95 latency and error rate missed
SLOs

SLO scorecard

ObjectiveTargetActualStatus
p95 latency< 3,000 ms3,195 msMissed
Error rate< 5%100%Missed
Tool failure rate< 2%0%Met
Average cost per task< $0.05$0.0000Met

What the data told me, and what I did

FindingMeal recommendations failed on every call

All 33 traced calls errored on both Gemini and GPT-4o-mini. The fallback switched providers, but the request itself failed in both places.

Action

Debug the request itself, since both providers failed on the same input, and add a curated fallback so people always get ideas.

Finding"Cost met" was a false positive

$0.0000 per task looked great, but only because failed calls cost nothing.

Action

An SLO only counts when its dependencies pass. Cost per successful task is the real metric.

FindingThe fallback is 7.5× slower

GPT-4o-mini averaged ~2,460 ms against ~327 ms for Gemini Flash, which pushed p95 past the 3-second target.

Action

Keep the fast model on the voice path, use the slower one only off the critical path, and add a timeout before falling back.

FindingVoice logging is used but not loved

8 uses, 17% positive, and "Didn't understand" as the top reason.

Action

Build a Hinglish meal-phrase eval set and score parsing accuracy on every prompt change.

FindingOnly one feature sends LLM traces

Every traced call came from meal recommendations, while voice logging and behaviour analysis showed usage but no traces.

Action

Instrument every AI feature the same way before adding new ones.

Closed beta with 4 active users, so the numbers point to problems rather than prove trends. The point was to build the measurement loop before growth, not after.

The evaluation loop

Ship a prompt version→ Trace every call→ Collect thumbs + reasons→ Check SLOs→ Compare versions→ Fix and re-run ↺

Next: an offline eval set of real meal phrases scored before every release, and LLM-as-judge scoring for meal-idea quality, so I can catch regressions before users do. The console already has an "AI version performance" view waiting for that data.

09How users benefit

What Priya gets out of it

TimeSeconds to log, not minutes

Voice or one-tap confirm replaces searching a database and guessing portions.

RelevanceFood she actually eats

Daliya, dal and roti are in the plan, with portions in bowls and rotis.

ResilienceA bad week isn't the end

Pause, rest days and treat days keep her in the habit through travel and illness.

GuidanceAlways knows what's next

Three meal ideas on demand, matched to mood, cuisine and allergies.

MotivationWins, not warnings

Progress leads with strengths and small streaks instead of red numbers.

TrustNo surprise changes

Every plan adjustment shows what changed and why.

10Success metrics

What I'm measuring

Targets for a 90-day beta, not results yet.

North starUsers with 5+ logged days a week, in week 4

It directly measures closing the disengagement gap.

  • Target 35% of signups
Input metricsIs logging easy and sticky?
  • Median time to log a meal < 10s
  • Meals logged via voice or quick confirm > 70%
  • Users who pause and come back > 50%
  • Meal ideas generated per WAU 2+ / week
  • Auto-adjusted plans kept, not reverted > 80%
GuardrailsIs it safe and reliable?
  • AI request failure rate, after fallback < 0.5%
  • Voice parse accuracy on eval set > 90%
  • Adjusted targets within safe bounds 100%
  • p95 AI response time < 3s
  • AI cost per monthly active user < $0.40
11Tradeoffs

Trade-offs I made

DecisionWhat I gave upWhy it was worth it
Consistency over precisionFittr-level gram accuracyAn approximate log every day beats a perfect log for one week.
Voice and plan-confirm, not photos firstThe wow factor of photo AIPhotos struggle with hidden fats and mixed dishes. Confirming a planned meal is faster and more accurate.
Pause instead of streak pressureThe urgency streaks createStreak anxiety is exactly what makes restarters quit for good.
Coach and Community deferredHuman accountability, Fittr's biggest strengthProve the solo habit loop first, then add humans where AI falls short.
Show every plan changeA cleaner, simpler screenChanging someone's diet without explanation breaks trust.
12Roadmap

What's next

Now

Make it dependable

  • Curated fallback when both AI providers fail
  • Auto Adjust checks logging completeness first
  • Keep logs and progress always in sync
Next

Close the gaps

  • Wearable sync: steps, sleep, Apple Health
  • Maintenance mode after goal weight
  • Hinglish voice evaluation set
Later

Add the human layer

  • Hybrid AI plus human coach tier
  • Small accountability circles, not public feeds
  • Family plans: one kitchen, many goals
13What I learned

What I learned

Measure before you scale

With four beta users, my SLO dashboard still showed a 100% error rate on meal ideas across both providers. Without that instrumentation, I would have blamed the users for not engaging.

Data needs context

"71% under target" looked like a reason to eat more. It was probably a reason to log more. AI that adjusts people's diets has to ask before it acts.

Empathy can be designed

Rest days, pause and a gentle greeting are small features, and they speak directly to the reason people quit.