Cleveland Kitchen - PoC Analysis

Date: February 10, 2026
Location: Cleveland, Ohio
Industry: Food Manufacturing - Fermented Foods (Kimchi, Pickles, Krauts)
ERP: WhereFour (inventory management) + QuickBooks

---

Company Overview

Cleveland Kitchen is a fermented foods manufacturer producing:

• Kimchi (classic, spicy, and variations)
• Krauts (sauerkraut with various flavors: caraway, curry, etc.)
• Pickles (dill, bread & butter)
• Salsas (fermented red and green)

Order Volume:

• Total: 500-600 orders/month
• EDI (SPS platform): ~90% of orders (automated via SPS)
• Manual email orders: ~10% (50-100 orders/month)

Current Team:

• Ben Bvandertill (Customer Service Manager) - oversees order flow
• Sam Buddenbaum (Operations/Management)
• Philippines outsourced team - manual order entry (fixed rate per head)

---

Why Cleveland Kitchen is a Great Fit

✅ Clear Pain Point: Manual Order Entry

• 50-100 manual email orders/month
• Processing time: ~10-15 minutes per order (entry + verification + logistics coordination)
• Outsourced to Philippines team (cost: fixed rate per head)
• Varied formats: PDF attachments (most), email body text, Excel files

✅ Messy SKU/UPC Mapping

• Major pain: UPC codes vary by customer (12-digit vs. 14-digit, leading/trailing zeros)
• Customer descriptions highly varied - no standardization
• Mapping project when onboarding SPS EDI took significant time
• Cross-reference table critical for matching customer IDs → Cleveland Kitchen SKUs

✅ Culturally Ready for Automation

• Already outsourcing order entry (less political complexity)
• Not replacing US jobs - enhancing Philippines team efficiency or reducing dependency
• Comfortable with AI/automation concepts

✅ Strategic Value: WhereFour Integration

• First WhereFour customer = opens up food distribution market
• WhereFour appears common in food/beverage industry
• Building integration unlocks future customers in similar verticals

✅ Local & Collaborative

• Cleveland-based = easy to visit, hands-on collaboration
• Willing to be “design partner” for R&D phase
• Open to flexible pricing/pilot structure

---

PoC Scope

IN SCOPE ✅

Phase 1: Email Extraction + Validation (Weeks 1-4)

1. Email Ingestion

   • Monitor shared customer service inbox
   • Triage: order vs. non-order emails
   • Handle PDF attachments, email body text, Excel files

2. Field Extraction

   • PO number (if present - some food service customers don’t use formal POs)
   • Customer identification (company name, contact)
   • NEW: Shipping address, billing address, location/warehouse (all distinct fields)
   • Line items: product description, quantity, price
   • NEW: Discount/allowance line items (negative pricing) - must be flagged visually
   • Requested delivery date (2-week lead time standard)
   • Special instructions
   • NOTE: Cisco orders have heavy contract/boilerplate language → filter noise, extract key fields only

3. UPC/SKU Normalization

   • Detect 12-digit vs. 14-digit UPCs
   • Handle leading/trailing zeros
   • Map customer product descriptions → Cleveland Kitchen master SKU list
   • Build cross-reference table over time (AI learns customer-specific terminology)

4. Pricing Validation

   • Check if customer/SKU combination exists in WhereFour pricing tiers
   • Flag for manual review if new pairing (first-time order for that combo)
   • Extract allowances/discounts as separate line items

5. “Same as Last Time” Orders

   • Detect reference to previous order (“same as last week”, “repeat order from 1/15”)
   • Extract reference order number or date
   • Human review required to pull previous order details

6. Human-in-the-Loop Validation

   • CSR or Philippines team reviews extracted data
   • Corrects any mismatches
   • Approves for submission to WhereFour

Phase 2: WhereFour Integration (Weeks 5-10) (Conditional on Phase 1 success)

1. API Connection to WhereFour

   • Product lookup endpoint
   • Customer lookup endpoint
   • Order creation endpoint
   • Pricing tier validation

2. Order Submission

   • Create order in WhereFour with validated data
   • Link to QuickBooks (existing WhereFour → QB API remains unchanged)
   • Trigger logistics flow (existing WhereFour automation)

3. Error Handling

   • NEVER block due to low inventory (Cleveland Kitchen manages MRP in Excel)
   • Flag pricing mismatches
   • Flag new customer/SKU pairings
   • Surface errors to CSR team for resolution

---

OUT OF SCOPE ❌

1. EDI orders via SPS - Already automated, ~90% of volume
2. MRP/inventory planning - Runs externally in Excel, not in WhereFour
3. Logistics scheduling - Handled by logistics team post-order entry
4. Production planning - Fermentation scheduling remains manual
5. QuickBooks integration - Existing WhereFour → QB API unchanged

---

Technical Requirements

ERP: WhereFour + QuickBooks

• WhereFour: Cloud-based inventory management software
• QuickBooks: Accounting system (connected via API)
• Current flow: WhereFour → QuickBooks (automated) → Logistics team
• Integration status: ⚠️ WhereFour integration does not exist yet - Cleveland Kitchen is first customer
• API documentation: To be provided by WhereFour or Cleveland Kitchen IT

Data Replication Needs

1. Products - erp_product_id, name, description, UPC (12 & 14-digit), category, pricing
2. Customers - customer_id, company_name, ship-to addresses, customer type (food service vs. retail)
3. Cross-Refs - customer UPC/description → Cleveland Kitchen master SKU mappings
4. Pricing Tiers - customer-level pricing agreements (built in WhereFour)
5. Allowances - discount SKUs (negative pricing) mapped to customer agreements

Mockoon Mock

• Status: ⚠️ Does not exist yet - Cleveland Kitchen is first WhereFour customer
• Action: Build new WhereFour Mockoon mock for dev/testing
• Endpoints needed:
  1. GET /products - product lookup with UPC variations
  2. GET /customers - customer lookup
  3. POST /orders - create order
  4. GET /pricing-tiers - customer-specific pricing validation
  5. (Optional) GET /order-history - for “same as last time” reference orders

---

Business Rules (Critical for Action Layer)

UPC/SKU Mapping

• Challenge: Customers use inconsistent UPC formats and descriptions
• 12-digit UPC: 850004476017 (most common)
• 14-digit UPC: 00850004476017 (adds leading zeros)
• Some customers drop leading zeros, others add trailing zeros
• Solution: Normalize all UPC variations, maintain cross-reference table per customer

Inventory Handling

• CRITICAL: Do NOT block orders due to low/no inventory
• Cleveland Kitchen’s MRP (Material Requirements Planning) runs externally in Excel
• WhereFour inventory data is not reliable for decision-making
• Orders must enter WhereFour regardless of stock status
• Production planning team handles fermentation scheduling (1-week lead time)

Pricing & Address Validation

• Pricing tiers built at customer level in WhereFour
• New customer/SKU combo = flag for manual review (CSR approves pricing)
• Some customers have “everyday low cost” agreements (e.g., Giant Eagle @ $20/unit vs. list $21.50)
• Allowances = discounts/rebates as separate line items (negative pricing)
  • Example: Order 6 SKUs → 6 allowance lines = 12 total lines
  • UI Requirement: Show visual chip/indicator (e.g., yellow badge “discount applied”) so reviewers don’t override
  • Reason: Trust ERP price over PDF price; surface the adjustment so humans understand what happened

Address Types (NEW from onsite)

• Shipping address = where products ship
• Billing address = different from shipping (for some customers)
• Location/Warehouse = customer’s distribution center (may differ from both shipping + billing)
• Implementation: Each address type becomes a separate flag in WhereFour lookup
• Validation: Must validate customer + shipping address combo in WhereFour (determines pricing tier)

Lead Time

• Standard request: 2 weeks advance notice
• Reality: Customers vary (some give less notice)
• Constraint: Fermentation takes ~1 week for most products
• Inventory target: 2-3 weeks on hand (but often hand-to-mouth)

“Same as Last Time” Orders

• Some customers order by reference: “Same as order from Jan 15” or “Repeat last week’s order”
• Requires: Lookup of previous order in WhereFour order history
• Human review: CSR confirms quantities, pricing haven’t changed
• Future enhancement: AI could auto-suggest previous order match

---

Open Questions / Clarifications Needed

RESOLVED ✅ (from March 5 onsite meeting)

1. Excel File Extraction ✅

   • Both .xlsx (Giant Eagle format) and .xls (Heinen’s legacy format) now supported
   • Fixed: openpyxl + xlrd fallback for binary .xls files
   • Demo blocked by this on 3/5 → fixed in staging the same day

2. Giant Eagle & Heinen’s Priority Suppliers ✅

   • Giant Eagle: Orders come in .xlsx format, ~15 lines per order, all at $20/unit (custom pricing)
   • Heinen’s: Orders come in .xls format, legacy binary Excel
   • Both are majority of manual orders (need real-world walkthrough video)
   • Plan: Produce demo video with actual extractions from both formats

3. Discount/Allowance Handling ✅

   • Cleveland Kitchen uses “custom line items” in WhereFour (not inventory SKUs)
   • Discounts appear as negative pricing line items in orders
   • Need to surface discount as visual chip/indicator in Ordermatic UI (so humans don’t feel need to override)
   • Example: Order at $20, WhereFour price is $21.50 → show visual “discount applied” indicator

4. Multiple Address Types ✅

   • Shipping address (where products ship)
   • Billing address (different for some customers)
   • Location/warehouse (customer’s DC, separate from shipping)
   • Need new flag: location field to be extracted + validated in WhereFour

5. Email Thread Management ✅

   • Important decision: Start without AI reading every email thread
   • Pick up initial PO only, ignore reply threads for now
   • Ben + Sam will make manual changes to orders in Ordermatic as needed
   • Reason: They change orders hourly, so thread-based AI would create chaos
   • Future enhancement: Thread skill for specific scenarios (e.g., “if discount is wrong, generate new quote”)

6. Auto-Submission to WhereFour ✅

   • Vision: If order has zero validation issues, auto-submit to WhereFour (no human required)
   • Philippines team only does QA review (5-30 sec “is this correct?” check)
   • See integration log in WhereFour to spot errors
   • Helps establish eval set for Phase 2 automation

STILL OPEN ❓

7. WhereFour API Documentation

   • REST endpoints available?
   • Authentication method?
   • Rate limits?
   • Webhook support for order status updates?
   • Status: Dan said he’ll be responsive; David to prepare specific API questions for them

8. Master Item List

   • Export from WhereFour: all SKUs with UPC variations, descriptions, pricing
   • Include allowance SKUs (discount line items)
   • Expected: Will be part of WhereFour sandbox access

9. Customer Cross-Reference Table

   • Existing mapping: customer UPC/description → Cleveland Kitchen SKU
   • Format: CSV or Excel export
   • Expected: Will provide along with 4-month historical data

10. Training/Historical Data

    • Screen recordings of manual order entry process (Cisco order example showed complex boilerplate)
    • 4 months of historical email orders for ROI analysis (to show time savings vs. cost)
    • Status: Dan agreed to send 4-month history (Jan-Feb selected to avoid October/November demand skew)

11. Order History API

    • For “same as last time” orders, can we query WhereFour order history via API?
    • Or does CSR manually look up in WhereFour UI?
    • Status: Deferred to Phase 2 (Phase 1 focuses on extraction only)

---

Pricing & ROI

Proposed Structure (from transcript)

Item	Cost	Notes
WhereFour Integration	$15,000 → $0	Waived as “design partner” (first WhereFour customer)
Monthly Platform Fee	$500/mo	Organization-level (not per-seat)
Per-Line-Item Fee	$0.40/line	Minimum $2.50 per order

Monthly Cost Estimate

• Volume: 50-100 orders/month
• Avg SKUs per order: 4-8 SKUs (+ allowances can double to 8-16 lines)
• Conservative estimate: 75 orders/month × 6 lines/order = 450 lines/month
• Line item cost: 450 × $0.40 = $180/month
• Platform fee: $500/month
• Total: ~$680/month or $8,160/year

ROI Comparison

• Current state: Philippines team (fixed rate per head) handling ~10-15 min/order
• Estimated time savings: If 50-80% of orders auto-validated → 5-10 hours/month saved
• Error reduction: Fewer SKU/UPC mapping errors, pricing mismatches
• Scalability: Volume can fluctuate without hiring/training overhead

R&D Phase Agreement (from transcript)

• Phase 1 (Extraction PoC): Ordermatic does R&D at no cost
• Cleveland Kitchen provides: NDA, sample orders, master item list, screen recordings
• Decision point: After Phase 1, Cleveland Kitchen decides:
  • If YES → Pay normal $15K integration fee + ongoing usage fees
  • If NO → No cost, part ways as friends (Ordermatic keeps WhereFour integration for future customers)
• Incentive if YES: 3 months free service (waives $5K implementation fee)

---

Implementation Plan

Phase 0: Discovery & Data Collection (Week 1)

• ✅ NDA signed
• ✅ Cleveland Kitchen sends 10-15 sample orders (clean + messy)
• ✅ Master item list export from WhereFour
• ✅ Customer cross-reference table
• ✅ Screen recordings of manual order entry process
• ✅ WhereFour API documentation (if available)

Phase 1: Email Extraction PoC (Weeks 2-4)

• ✅ Set up Cleveland Kitchen PoC org in Ordermatic
• ✅ Configure email forwarding (customer service inbox → Ordermatic)
• ✅ Build extraction logic for Cleveland Kitchen-specific formats
  • ✅ Excel extraction (both .xlsx Giant Eagle + .xls Heinen’s legacy format)
  • Extract boilerplate filtering (Cisco orders show heavy contract language)
  • Email threading: Pick initial PO only, ignore reply threads
  • Discount/allowance detection with visual chip indicator
  • NEW: Extract all address types (shipping, billing, location)
• ✅ Train AI on UPC/SKU mapping variations
• ✅ Test extraction accuracy on sample orders (target: >90% accuracy)
• NEW: Create video demo with Giant Eagle + Heinen’s real extractions
• Deliverable: Extraction demo + video walkthrough for Ben + Philippines team

Phase 2: WhereFour Integration (Weeks 5-10) (Conditional on Phase 1 success)

• Build Mockoon WhereFour mock for dev/testing
• Implement WhereFour API endpoints (product, customer, order, pricing tiers)
• Add business rule validations (pricing, allowances, new customer/SKU flagging, address validation)
• CRITICAL: Disable inventory blocking (per Cleveland Kitchen requirement)
• NEW: Auto-submission workflow
  • If order has zero validation issues → auto-submit to WhereFour (no human required)
  • Philippines team reviews only problematic orders (5-30 sec QA check)
  • Log all submissions in WhereFour integration log (for error diagnosis)
  • Helps establish eval set for Phase 3+ automation
• Error handling & reporting (surface errors to CSR team in Ordermatic)
• UAT with Ben + Philippines team

Phase 3: Production Deployment (Weeks 11-12)

• Deploy to production
• Monitor extraction accuracy
• Philippines team validates orders in Ordermatic → submits to WhereFour
• Feedback loop: Capture missed mappings, build cross-reference table over time
• Iterate on AI extraction based on real-world edge cases

Phase 4: Optimization (Ongoing)

• Build “same as last time” order reference lookup
• Enhance allowance detection
• Automate pricing tier validation
• Reduce human-in-loop validation to <20% of orders
• Email Thread Skill (future enhancement, not Phase 1)
  • When reply email arrives: AI listens for specific order modifications (e.g., “change DC”, “discount is wrong”)
  • Option 1: Generate new quote with proposed changes, send to inbox
  • Option 2: Auto-update order in Ordermatic for CSR approval
  • Status: Not in Phase 1 (too risky given hourly changes); evaluate in Phase 4 after understanding CK’s email patterns

Long-term Migration Path (CK’s Strategic Decision)

• Cleveland Kitchen is evaluating Prisma - a custom ERP for food manufacturers (early-stage)
• Implication for Ordermatic: WhereFour integration may have short shelf life
• Opportunity: David should connect with Prisma founder (contact: Dan will intro)
  • Prisma needs order entry automation too
  • Ordermatic could become consultant/integration partner for new ERP
  • Longer-term: Migrate CK from WhereFour → Prisma, keep Ordermatic extraction + validation layer
• API Documentation: David to build custom WhereFour API docs (better than official) → will transfer to Prisma integration later

---

Success Criteria

Phase 1 Success (Week 4)

• ✅ 90%+ extraction accuracy on sample orders
• ✅ Correct UPC normalization (12-digit ↔ 14-digit)
• ✅ Customer/product matching functional
• ✅ Allowances detected as separate line items
• ✅ Ben + Sam approve quality of extracted data

Phase 2 Success (Week 10)

• ✅ Orders successfully created in WhereFour
• ✅ Pricing validation working (flags new customer/SKU combos)
• ✅ Inventory blocking disabled (orders enter regardless of stock)
• ✅ Error rate <2% (measured against manual entry baseline)
• ✅ Philippines team reports time savings

Phase 3 Success (Week 12)

• ✅ 60-80% of orders require minimal human editing
• ✅ CSR team trusts system enough to reduce validation time
• ✅ Cross-reference table growing automatically (AI learns customer terminology)
• ✅ Positive feedback from Ben + Philippines team
• ✅ Cleveland Kitchen decides to move forward with paid agreement

---

Risks & Mitigation

Risk 1: UPC/SKU Mapping Complexity

• Impact: HIGH - Core challenge for Cleveland Kitchen
• Mitigation:
  • Build robust UPC normalization logic (12 ↔ 14 digit)
  • Maintain per-customer cross-reference table
  • Human-in-loop for first occurrence of new customer/SKU combo
  • AI learns mappings over time

Risk 2: WhereFour API Limitations

• Impact: MEDIUM - Unknown until API documentation reviewed
• Mitigation:
  • Early API exploration before Phase 2 commitment
  • Fallback: If API insufficient, evaluate alternative integration methods (ODBC, CSV export/import)

Risk 3: “Same as Last Time” Orders

• Impact: MEDIUM - Require order history lookup
• Mitigation:
  • Phase 1: Flag for human review
  • Phase 4: Build automated lookup if WhereFour API supports order history queries

Risk 4: Allowance Detection

• Impact: MEDIUM - Doubles line count, affects pricing
• Mitigation:
  • Train AI on allowance keywords (“discount”, “rebate”, “allowance”, “promo”)
  • Validate against known allowance SKUs in WhereFour
  • Human review for first occurrence per customer

Risk 5: Philippines Team Adoption

• Impact: LOW-MEDIUM - Team needs to trust system
• Mitigation:
  • Involve Ben (Customer Service Manager) early and often
  • Provide training on Ordermatic UI
  • Emphasize: Not job replacement, task simplification
  • Show time savings on repetitive orders

---

Competitive Landscape

Existing Cleveland Kitchen Solutions

• EDI (SPS platform): Handles 90% of volume (large customers)
  • Setup cost: Unknown (already implemented)
  • Per-transaction fee: Unknown
  • Limitation: Requires customer EDI capability (smaller customers can’t do this)
• Philippines outsourcing: Fixed rate per head
  • Limitation: Manual entry errors (UPC mapping), time-consuming, not scalable

Why Ordermatic Wins

• EDI requires customer buy-in; Ordermatic works with any email format
• More flexible than outsourcing (no offshore lag, instant validation)
• Scales with business growth (seasonal fluctuations)
• Reduces SKU mapping errors (AI learns customer-specific terminology)
• Strategic value: Opens up WhereFour integration for future food/beverage customers

---

Next Steps (Action Items)

Immediate (This Week)

1. ✅ NDA - Sam to send NDA to Ordermatic
2. ✅ Sample Orders - Ben to send 10-15 email orders (variety of formats)
   • ✅ Clean PDFs (Four Seasons, Misfits Market, Rainforest, SirNA Sons, Sysco)
   • ✅ Excel files (Giant Eagle @ 15 lines/order, Heinen’s legacy .xls format)
   • ✅ Email body text orders
   • ✅ Cisco order (with heavy contract boilerplate - good test case)
   • Include at least one “same as last time” reference order
3. ⏳ Master Item List - Export from WhereFour (all SKUs, UPCs, descriptions, pricing)
4. ⏳ Cross-Reference Table - Existing customer UPC/description mappings (if available)
5. ⏳ Training Materials - Screen recordings of manual order entry process
6. NEW: 4-Month Historical Data - Dan will send Jan-Feb email order records (for ROI analysis)
   • Exclude Oct-Nov (anomalous demand, would skew weighted average)
   • Used to calculate: time savings vs. cost if using Ordermatic
7. NEW: WhereFour Sandbox Access - Dan to provide sandbox account (currently $150-250/mo for one user)
   • David will explore API, prepare list of unanswerable questions for WhereFour support
8. NEW: Email Configuration - They use Outlook with Gmail forwarding (some team members)
   • Standard Gmail servers route mail through, so affordable integration works fine
   • Alternative: OAuth app for Outlook (gives more control, but not needed for Phase 1)

Week 2-4: Extraction PoC

6. ⚙️ Ordermatic builds extraction logic for Cleveland Kitchen formats
7. ⚙️ Test extraction on sample orders, measure accuracy
8. ⚙️ Schedule follow-up demo to show extraction results
9. ⚙️ Cleveland Kitchen decides: Proceed to Phase 2 or not?

Week 5-10: WhereFour Integration (if Phase 1 approved)

10. ⚙️ Obtain WhereFour API documentation
11. ⚙️ Build Mockoon WhereFour mock
12. ⚙️ Implement order creation, pricing validation
13. ⚙️ UAT with Ben + Philippines team
14. ⚙️ Deploy to production

Ongoing

15. 🔄 Monitor extraction accuracy, iterate on AI training
16. 🔄 Build cross-reference table over time
17. 🔄 Expand to additional Cleveland Kitchen customers (if successful)

---

Conclusion

Cleveland Kitchen is a strong fit for Ordermatic’s first WhereFour integration:

• Clear pain point: Manual email order entry (50-100/month)
• Messy SKU/UPC mapping = high-value AI learning opportunity
• Culturally ready: Already outsourcing, open to automation
• Strategic value: WhereFour integration opens food/beverage distribution market
• Local collaboration: Cleveland-based, easy to visit and support
• Flexible partnership: Willing to be “design partner” for R&D phase

Recommendation: Proceed with Phase 1 Extraction PoC (R&D at no cost). If successful, negotiate Phase 2 WhereFour integration with Cleveland Kitchen as design partner (waive $15K fee, offer 3 months free service).

Strategic Priority: WhereFour integration = unlock future customers in food/beverage vertical (high-value market opportunity).

---

March 5, 2026 Onsite Meeting - Summary & Outcomes

Demo Status

• ❌ Excel extraction failed during demo (Giant Eagle .xlsx file didn’t process)
• Root cause: Legacy xlrd dependency removed in pandas 2.0+ (openpyxl used for .xlsx only)
• Fixed same day: Added xlrd fallback for binary .xls files, tested both formats
• ✅ All other extractions worked (PDFs, email text, Cisco order)
• ✅ Demo showed pricing validation, address lookup, allowance detection

Key Learnings

1. Excel Files Are Critical

   • Giant Eagle orders: .xlsx format, ~15 lines/order, consistent format
   • Heinen’s orders: .xls legacy binary format (requires xlrd)
   • Both are highest-priority suppliers (frequent manual orders)

2. Discount/Allowance Handling

   • Custom line items in WhereFour (not inventory SKUs)
   • Appear as negative pricing in orders
   • UX requirement: Visual chip indicator (yellow badge) to show “discount applied”
   • Reason: Reviewers shouldn’t feel need to override if ERP already has correct price

3. Email Threading

   • Unanimous decision: Don’t AI-process reply threads in Phase 1
   • Pick initial PO only → manual order adjustments for hourly changes
   • Reason: Production scheduling changes constantly; thread-based AI would create chaos
   • Future: Email thread skill (Phase 4) for specific scenarios

4. Address Types

   • Three distinct fields: Shipping, Billing, Location/Warehouse
   • Pricing may vary by location (e.g., different DC = different pricing tier)
   • Implementation: Extract all three + validate shipping/location in WhereFour

5. Auto-Submission Vision

   • If order has zero validation issues → auto-submit to WhereFour
   • Philippines team: QA only (5-30 sec “is this correct?” check)
   • Builds eval set for Phase 3+ automation
   • Integration log in WhereFour for error diagnosis

6. Long-term ERP Migration (Strategic Opportunity)

   • Cleveland Kitchen evaluating “Prisma” (custom ERP for food manufacturers)
   • WhereFour integration may have limited shelf life (1-2 years)
   • Opportunity: Ordermatic could integrate Prisma later; David should connect with founder

Data Commitments Received

• ✅ 4-month historical email order data (Jan-Feb selected for ROI analysis)
• ⏳ WhereFour sandbox access (for API exploration)
• ⏳ Master item list + customer cross-reference table
• ⏳ Training materials (screen recordings of manual entry)

Next Milestone

• David to produce video demo with Giant Eagle + Heinen’s real extractions (by end of week)
• Video showcases both Excel formats working smoothly
• Present to Ben + Philippines team for feedback before Phase 2 commitment