BACCHUS

Resolves whether multiple signals are talking about the same restaurant. When a permit, a press mention, and a LinkedIn post all reference slightly different names at the same address, the Causal Engine figures out they're the same place — and builds a timeline of where they are in the opening process.

Projects when a restaurant will open based on permit timelines specific to each metro. A Houston TABC application typically runs 120 days to opening; a Manhattan SLA permit runs 150. The Temporal Engine uses this to estimate your outreach windows.

Stress-tests every score by looking for contradictions. High luxury score but no indication the concept has committed to the market window? Red flag. Strong neighborhood proxy but zero operational signal to back it up? Red flag. These flags reduce confidence and can push an account back to MONITOR.

Takes the fit score, the confidence level, and any adversarial flags, then makes the final call: SURFACE, MONITOR, or EXCLUDE. This is what decides whether something lands on your Dashboard.

Runs every night to catch data problems — stale records that haven't been updated in 30 days, signals that contradict each other (a closure notice for an account still marked active), or duplicate entries for the same restaurant. Keeps the signal clean.

The score panel shows the AURORA decision, fit score, and a breakdown of all seven scoring dimensions. The Reputation & Market card shows Google rating and review count, press score, luxury proximity index, and area affluence (ZIP income index). The Route Planner shows nearby accounts within 10 miles — useful for clustering visits on a market trip.

The page header shows prestige badges (Michelin stars, James Beard, Forbes, Relais & Châteaux), average check range, seat count, and weekly covers.

The Beverage Intel panel shows the depth of the champagne and sparkling wine program — from None to Deep. This is a proxy for a guest base that expects high-end service pairing. You can edit it inline when you have first-hand knowledge.

The Caviar on Menu panel lists every species confirmed on the menu with origin, supplier, price point, and presentation. If a competitor product is already on the menu, it appears here.

Product Recommendations are generated automatically based on concept type, menu profile, and tasting history. An omakase gets an Osetra bump recommendation. A steakhouse gets a Beluga Hybrid pairing. Previous tasting preferences override the default.

All contacts sorted by priority: Executive Chef, Chef de Cuisine, Sous Chef, General Manager, Beverage Director, Sommelier, Purchasing Manager, Owner, Operating Partner. Each contact shows verified status, current/former role, phone, email, and LinkedIn. Verified contacts have been corroborated by at least two independent sources.

If this account has an active pipeline opportunity, you'll see the current stage, assigned rep, sample drop history, and tasting records. Below that is the activity log — a chronological record of every call, email, visit, note, and follow-up. Log new activities inline: select the type, write a subject and notes, record an outcome, and optionally schedule the next follow-up.

A unified timeline of all activity: logged calls and visits, sample drops (product, quantity, date), and tastings (chef reaction, preferred species, outcome). Sorted newest-first. This tab is the full history — the Pipeline tab shows the most recent; Activity shows everything.

Monday

• ·Open Dashboard — review this week's top 15 SURFACE accounts
• ·Open Openings feed — check for new accounts in the red or amber zone
• ·Note any new accounts that weren't on the list last week
• ·Check Events calendar for anything triggering in the next 30 days
• ·Prioritize outreach: Tier 1 first, then Tier 2, then amber-zone pre-openings

Tue – Thu

• ·Execute outreach on surfaced accounts
• ·Before each call: open the account detail page, review the Intelligence tab for beverage depth and product recommendations
• ·Log every call, visit, and sample drop in Pipeline immediately after
• ·After a tasting, log chef feedback: reaction, preferred species, objections — don't wait until Friday
• ·Use Route Planner when in the field to pick up nearby stops

Friday

• ·Update Pipeline — move any accounts that progressed this week
• ·Check Accounts view for MONITOR accounts that may have received new signals
• ·Directors: review Reports for conversion rates and pipeline health
• ·Directors: review Territory Health to identify stalled pipeline and coverage gaps

Why is a restaurant I know is great sitting in MONITOR?

Score and confidence both have to clear the threshold. The score might be strong but if we only have one or two signals on record, confidence hasn't reached the gate. It will tip over when the next batch of data comes in — usually within a week.

Can I manually add an account?

Not yet. Right now all accounts enter the system through the ingestion pipeline. If you know of a venue that should be tracked, flag it to your admin — they can adjust the ingestion coverage.

What does it mean when AURORA says "insufficient chef signal"?

The system hasn't identified a named executive chef for the concept. This triggers an adversarial flag because culinary vision without an identified leader is harder to validate. It doesn't mean the account is dead — it means you should try to confirm who's in the kitchen before the tasting call.

Why does a restaurant near me show "EXCLUDE"?

EXCLUDE usually means a contradiction was detected — a concept that looks luxury on paper but with signals that don't hold up (e.g., strong neighborhood score driven purely by zip code, no operational corroboration). It can also trigger if a closure signal came in. Check the account detail page — the adversarial flags section explains the specific reason.

How often does the data update?

Permit data refreshes every Monday at 6am. Press and personnel signals update twice daily. Menu data updates every two weeks. Scores are recomputed whenever a new signal arrives for an account.

The projected opening date looks wrong — what do I do?

Opening projections are based on historical permit-to-opening timelines for each metro. They're estimates, not guarantees. If you have boots-on-the-ground intel that contradicts the projection, log a note in the Pipeline. Your direct knowledge always beats the model.

What do I do with a Tier 1 account vs a Tier 3?

Tier 1 (90+) is your highest-conviction pitch — bring your full arsenal, request a private tasting, loop in your director. Tier 3 (75–79) is still worth pursuing but with a lighter initial touch — a sample drop and a follow-up call first.

What are the Product Recommendations on the Intelligence tab based on?

Recommendations are generated from three inputs: (1) concept type — omakase and Japanese concepts get Osetra recommendations, steakhouses get Beluga Hybrid, raw bars get Siberian; (2) menu content — existing luxury items on the menu inform cross-sell opportunities; (3) tasting history — if a chef has tasted and shown preference for a specific species, that overrides the default. The recommendation engine runs fresh on every page load.

Why does the Google rating show on some accounts but not others?

Google Places enrichment runs in batches from Settings. If an account doesn't have a rating, it either hasn't been enriched yet or Google Places couldn't find a match. Go to Settings → Data Enrichment and run "Enrich Next 20" to process the next batch. Coverage should reach 90%+ after a few runs.

How do I share the Openings list with my team?

Use the "Copy List" button at the top of the Openings page. It copies a formatted plain-text list of the current filtered view — account names, locations, open dates, and AURORA tiers — ready to paste into Slack, email, or a weekly briefing doc.