Maximizing AI Insights: How to Add Bottles for Best Results
Get professional-grade wine analysis by providing the right information
Introduction: Intelligence-First Design
CollectorCellar.ai uses an intelligence-first approach to wine analysis. Unlike tools that rely on rigid constraints or databases, our AI acts as a Master Sommelier—analyzing your specific bottles using professional wine knowledge.
But like any sommelier, the quality of advice depends on the information provided.
This guide explains which fields matter most, how purchase context improves accuracy, and what AI insights you can expect based on your subscription tier.
The AI-Enhanced Fields ✨
When adding or editing a bottle, you'll notice certain fields marked with a burgundy background and a sparkle icon (✨). These are AI-enhanced fields that directly influence the quality of your AI insights.
Required Fields (marked with *)
* Label Name ✨
The specific wine name on the label.
Examples:
- Château Margaux
- Opus One
- Sassicaia
Why it matters: Different wines from the same producer have dramatically different profiles. "Reserve" vs "Estate" vs "Grand Vin" tells AI everything about intended quality and aging potential.
* Producer/Winery ✨
The producer or estate name.
Examples:
- Domaine de la Romanée-Conti
- Louis M. Martini
- Penfolds
Why it matters: Producer reputation and historical style inform AI about quality level, winemaking philosophy, and typical aging curves. A 2015 Cab from a legendary producer ages differently than one from a new winery.
* Vintage ✨
The harvest year—or check "Non-Vintage (NV)" for Champagne, Port, or blended wines without a vintage year.
Examples:
- 2015
- 2010
- NV (Non-Vintage)
Why it matters: Vintage determines current age, which is critical for drink window calculations. For NV wines, AI uses your purchase date to estimate effective age.
* Varietal ✨
The grape variety or blend.
Examples:
- Cabernet Sauvignon
- Pinot Noir
- Bordeaux Blend
- Champagne Blend
Why it matters: Different varietals age at different rates. Cabernet can age 20+ years; Pinot Noir typically peaks earlier. Blends require understanding of component proportions.
💡 Pro Tip: Capture Blend Details for Better Insights
When your bottle is a blend, specify the actual grape varieties instead of generic style terms. This dramatically improves AI accuracy.
Generic (Less Accurate):
- Rosé
- Red Blend
- White Blend
Specific (More Accurate):
- Hondarrabi Beltza (red), Hondarrabi Zuri (white)
- Cabernet Sauvignon, Merlot, Cabernet Franc
- Grenache, Syrah, Mourvèdre
- 60% Cabernet Sauvignon, 30% Merlot, 10% Cabernet Franc
Real example: A Basque Txakoli labeled as "Rosé" produced generic insights. When updated to "Hondarrabi Beltza (red), Hondarrabi Zuri (white)," AI correctly identified it as a traditional co-fermented Txakoli with ephemeral character, narrow drink window (2023-2026), and appropriate coastal-style pairings.
Why this matters: AI uses specific grape characteristics to reason about structure, aging potential, and pairing compatibility. "Rosé" is a color, not a varietal—it tells AI nothing about acidity, tannin, or regional style. The actual blend reveals everything.
If you don't know the blend composition, generic terms are fine. But when available (check the label, producer website, or wine databases), capturing the real blend unlocks significantly more accurate insights.
Recommended Optional Fields
Format ✨
Bottle size (375ml, 750ml, 1.5L, etc.)
Why it matters: Larger formats age slower due to lower oxygen-to-wine ratios. A Magnum (1.5L) of 2010 Bordeaux will peak later than a standard 750ml bottle.
Alcohol % ✨
The alcohol by volume percentage.
Why it matters: Higher alcohol affects body, intensity, and pairing recommendations. A 15.5% Zinfandel pairs differently than a 12.5% Pinot Noir.
Region/Appellation ✨
The wine's origin (country, region, or appellation).
Examples:
- Napa Valley, California
- Burgundy, France
- Barossa Valley, Australia
Why it matters: Regional characteristics influence style. Napa Cabernet is typically bolder than Bordeaux; Burgundy Pinot is more elegant than Oregon Pinot.
Subregion / AVA ✨
More specific origin (AVA, commune, vineyard).
Examples:
- Oakville AVA
- Pauillac
- Dundee Hills
Why it matters: Micro-climates and terroir affect quality and aging. Oakville Cabernet has different tannin structure than Rutherford Cabernet.
Purchase Price ✨
What you paid for the bottle.
Why it matters: Price correlates with quality, aging potential, and typical drink windows. A $200 Bordeaux is built to age longer than a $30 bottle.
Purchase Date ✨
When you bought the bottle.
Why it matters: Critical for Non-Vintage wines and helps AI understand how long the wine has been in your cellar vs. the producer's warehouse. A 2010 wine purchased in 2012 has aged differently than one purchased in 2024.
Non-AI Fields (Still Useful)
These fields don't influence AI analysis but help with organization and personal tracking:
- Currency - For purchase price tracking
- Open Date - Record when you drank it
- Preferred Open Date - Your planned opening target
- Storage Location - Where it's stored in your cellar
- Personal Tasting Notes - Your own impressions after tasting
- Rating (Basic tier and above) - Your star rating
What AI Insights You'll Receive
AI insights vary by subscription tier:
Enthusiast Tier
- AI Tasting Notes - Professional tasting profile including appearance, nose, palate, and finish
- AI Food Pairings - Specific dishes matched to the wine's structure
- 40 AI enrichments per month
Connoisseur Tier
Everything in Enthusiast, plus:
- AI Drink Window Estimates - When the wine will be at peak quality
- 80 AI enrichments per month
Cellar Pro Tier
Everything in Connoisseur, plus:
- Advanced Analytics - Deep insights into your collection
- 120 AI enrichments per month
Note: All AI tiers provide age-aware analysis based on current bottle age, not just vintage year.
The Intelligence-First Philosophy
CollectorCellar.ai doesn't use prescriptive rules like "Cabernet = ribeye."
Instead, it analyzes each bottle individually:
Traditional approach:
"Cabernet Sauvignon → automatically suggest ribeye"
Our intelligence-first approach:
"This is a 2015 Napa Cabernet from a quality producer at 14.8% alcohol. Consider its body, tannin structure, current age, and price point. Match wine weight to food weight. For inspiration, consider Italian cuisine with beef..."
This produces diverse, wine-specific pairings like:
- Osso Buco with gremolata
- Beef brasato with polenta
- Grilled ribeye with rosemary butter
- Mushroom risotto with aged Parmigiano-Reggiano
Tips for Best Results
1. Fill in as many AI-enhanced fields as possible
The more context AI has, the more accurate the analysis.
2. Be specific with names
"Reserve Cabernet" is better than "Cabernet"
"Opus One" is better than "Opus"
3. Use purchase data for NV wines
Non-Vintage Champagne and Port benefit greatly from purchase dates, which AI uses to estimate effective age.
4. Update bottles over time
If you try a bottle before finishing it, add your personal notes. When you re-generate AI insights later, they'll reflect the current age.
5. Re-generate insights as bottles age
AI drink windows are calculated based on current age. A 5-year-old wine analyzed again at 10 years will get updated recommendations.
6. Trust the intelligence
Our AI is trained to think like a Master Sommelier. If it suggests an unexpected pairing, there's likely a structural reason (acid, tannin, body, flavor intensity).
Example: Properly Adding a Bottle
Let's add a bottle step-by-step:
Bottle: 2015 Caymus Cabernet Sauvignon, Napa Valley
Required Fields:
- Label Name: Caymus Cabernet Sauvignon
- Producer: Caymus Vineyards
- Vintage: 2015
- Varietal: Cabernet Sauvignon
Recommended Optional Fields:
- Format: 750ml
- Alcohol %: 14.9
- Region: Napa Valley, California
- Subregion: (leave blank if unknown)
- Purchase Price: 85.00
- Currency: USD
- Purchase Date: 2023-12-15
Result:
AI will analyze this as a high-quality, age-worthy Napa Cab from a reputable producer, currently 11 years old, purchased recently but aged significantly. Expect rich, structured pairings and a drink window that accounts for its current maturity.
Validation: Real-World Accuracy
We validated our intelligence-first approach with many other AI models analyzing hundreds of diverse wines for example:
- 1971 Louis M. Martini Cabernet - "Highly accurate for legendary producer"
- NV Champagne - "Exceptionally well-suited pairings"
- 2008 Port - "Exceptionally reliable and accurate"
- 2022 Rosé - "Highly accurate" (drink now validated)
Overall validation: Hundreds of wines rated professional-grade
This confirms that our AI provides sommelier-level analysis when given proper information.
Conclusion: Information Enables Intelligence
CollectorCellar.ai's AI is only as smart as the information it receives.
By understanding which fields matter and why, you transform AI from a guessing tool into a professional wine advisor—one that considers your specific bottles, cellar diversity, and aging timelines.
Fill in AI-enhanced fields (✨), provide purchase context, and trust the intelligence.
That's how you get the most value from AI insights.
Ready to optimize your cellar?