How Amino AI Solves the Indian Protein Puzzle.
Generic western calorie trackers count crude nitrogen in salads and chicken breasts. They fail Indian households because Indian meals are built on complex grain, pulse, and dairy matrices. Here is how Amino AI demystifies desi protein with biochemical accuracy.
Download Android App (0.95 MB)The 4 Great Indian Protein Myths
Over 73% of Indian diets are protein-deficient according to ICMR data. The biggest reason is widespread dietary misinformation about what actually delivers usable protein.
I had a bowl of Dal, my protein is sorted.
The Reality: Raw dry pulses have ~22–24% protein. But home-cooked dal is 80–85% water. A standard 100ml bowl of thin watery dal contains only 12–15g of raw pulse equivalent — yielding just 2.5g–3.8g of protein, not the 15g people assume.
Smart volumetric dilution factors calculate exact raw-dal equivalents based on bowl size and gravy density.
Paneer has all the protein I need.
The Reality: A typical home-cooked paneer serving is only 40–50g (3–4 small cubes), delivering around 7–9g of protein. To hit even a 60g daily target from paneer alone, you would need to eat 350–400g — an unrealistic and expensive amount.
Amino AI shows you the exact amino acid profile and protein grams per katori of paneer — so you know precisely how much it contributes toward your daily 9-EAA completeness.
Soya Chaap & Soybeans are protein kings.
The Reality: Pure defatted TVP chunks have 52% protein. But street soya chaap is formulated with 60–70% refined maida and only 30% soy flour — reducing protein to under 8g per serving, served in oily gravies.
Differentiates defatted TVP chunks from commercial maida-heavy chaap formulations using IFCT 2017 entries.
I eat non-veg once a day, so I'm 100% fine.
The Reality: 2 boiled eggs give 13g protein. A 100g chicken curry portion gives 25–28g. For someone needing 70–85g daily, that single non-veg dinner still leaves a 35–50g deficit across breakfast and lunch.
Evaluates all-day pacing across every meal, unlocking hidden staple grain and dairy protein throughout the day.
9 Essential Amino Acids & Liebig's Law
Why dismissing Roti and Rice as “pure carbs” is a nutritional mistake, and how mutual supplementation unlocks complete muscle protein synthesis.
Liebig's Law of the Minimum — The Wooden Barrel
WHO / FAO (2007) Reference Model
Your body uses 20 amino acids to build muscle and repair enzymes. 11 can be made by your liver. But 9 Essential Amino Acids (EAAs) must come directly from food: Histidine, Isoleucine, Leucine, Lysine, Methionine (+Cys), Phenylalanine (+Tyr), Threonine, Tryptophan, and Valine.
Grains (Roti, Rice, Oats)
Wheat has ~11–12% protein. Rich in Methionine, but limited in Lysine (~35–45% of WHO requirement). Eaten alone, muscle synthesis stalls.
Pulses (Dal, Chana, Rajma)
Legumes are rich in Lysine, but deficient in sulfur amino acids like Methionine. Alone, up to 40% of protein is oxidized as urea.
The Thali Mutual Synthesis
When you pair Dal + Rice or Roti + Curd, amino profiles cross-supplement perfectly — achieving 100% biological value protein synthesis.
The 24-Hour Bodily Amino Acid Pool:
You don't need grain and pulse in the same mouthful. Your body maintains a circulating plasma amino acid pool for 12–24 hours. Amino AI tracks your full day's cumulative intake to ensure all 9 EAAs reach sufficiency by bedtime.
Zero-Hallucination Vision Scanning Engine
Why we strictly forbid AI from inventing nutrition numbers, and how our 4-step deterministic pipeline works.
Photo Capture
User snaps a home-cooked dish or selects a meal from the log. Fast camera pipeline works offline and online. Single-dish photos only — no full thali at once.
Vision AI — Identification Only
Multimodal vision identifies what dish it is and visual cues for portion estimation. Vision LLMs NEVER generate nutrition figures. They are strictly classifiers.
IFCT 2017 Deterministic Lookup
System queries the ICMR-NIN database for exact nutrient & 9-EAA values per 100g cooked state. No model inference — this is a fixed table lookup, immune to hallucination.
Portion Scaling
User confirms portion via Katori steppers (100ml, 180ml, 250ml) or piece counts. The scaler multiplies IFCT values by the confirmed volume to output exact gram values.
The Golden Rule: AI identifies. IFCT calculates. These two roles never cross. Any tracker that lets an LLM invent nutrition numbers is giving you statistically plausible fiction, not biochemical fact.
Gravy Dilution & Household Measure Logic
How Amino AI bridges the gap between chaotic Indian dinner tables and scientific precision — without asking you to weigh hot dal on a kitchen scale.
The 3 Standard Katoris
Calibrated against ICMR-NIN household dietary survey standards.
Dilution Equivalent Math
For liquid gravies, curries, and dals, the backend converts volume to raw pulse dry mass:
Eliminates the friction of weighing wet hot food on a scale.
Bounded Approximations
Trace spices (haldi, jeera), tempering oils (tadka), and sub-0.05g amino fractions have negligible metabolic impact on muscle protein synthesis.
We mathematically bound these variables without false lab-scale claims — giving you effortless speed with 98%+ macro integrity.
Personalized RDA Targets & Streak Grace Engine
Why rigid 100% daily targets cause tracking burnout, and how our calibrated streak criteria builds lasting habits.
Clinical Multipliers by Activity & Goal
Protein targets are dynamically computed from body weight (kg) and lifestyle demands:
Streak Baseline Target & Freeze Shields
Most fitness apps punish users with a broken streak if they hit 95% instead of 100%. Amino AI calculates a Personalized Streak Target locked at your baseline clinical RDA minimum — not your full aspirational goal.
Missed an unexpected travel day or a wedding feast? 1-Tap Streak Freeze Shields preserve your momentum, and smart recovery algorithms reactivate streaks once you log the next day.
Recipe Builder & Nutrition Intelligence
How Amino AI understands custom home cooking recipes and calculates real-time gap closer solutions so you end every day at complete protein.
Multi-Ingredient Recipe Architecture
Indian mothers rarely cook single ingredients. They mix palak with paneer, or combine 3 dals into a panchmel dal. Amino AI handles all of this:
- →Add multiple raw ingredients with IFCT 2017 lookups
- →System calculates moisture yield retention factors automatically
- →Aggregates individual amino matrices into an accurate per-katori score
- →Save recipes for 1-tap logging across future meals
Daily Nutrition Score (0–100) & Gap Closers
At the end of each day, our analytics engine evaluates:
- →Composite Score: Weighted balance of total protein grams (60%) + 9-EAA completeness (40%).
- →Gap Closer Engine: If dinner was deficient in Lysine or Methionine, the app suggests exact desi pairings (1 katori curd or 30g roasted chana) to bridge the gap before sleep.
Start Tracking Your 9 Amino Acids Today.
No food scales at the dinner table. No calorie anxiety. Verified ICMR-NIN science, frictionless katori portions, and complete protein synthesis — every day.