🔒 Internal Competitive Edge
ข้อมูลลับเกี่ยวกับ structural advantages — ห้ามเผยแพร่ภายนอก
Confidential information about structural advantages. Do not share externally.
Purpose
ส่วนนี้มีไว้เพื่อให้ทีมใช้งานข้างใน — ไม่ใช่ marketing เราไม่พูดเรื่องนี้กับลูกค้าเพราะ:
- (1) เป็น commercial partner terms ไม่ใช่ public claim
- (2) ผิดข้อตกลง partner ถ้า disclose
- (3) ถ้าลูกค้ารู้ว่าเราใช้ free credit จะทำให้ perceived value ตก
- (4) clients จ้างเราเพราะคนของเรา ไม่ใช่เพราะ supplier credit
This section describes structural advantages Go Digit has over any Thai competitor that we use internally to deliver at higher speed and quality — but we do NOT market publicly because:
- They are commercial partner terms, not public claims
- Disclosure could breach partner agreements
- Mentioning them lowers perceived value (clients think "oh, you're using free credits")
- They are structural, not features — clients should hire us for OUR work, not a supplier's credit
💎 The Asset: $200K USD AWS Partner Credit (Annual)
นี่คือ structural advantage ที่ไม่มีคู่แข่งในไทยมี — $200K USD/ปี ≈ ฿7 ล้าน/ปี ใน AWS credit ใช้ run Claude Opus, Sonnet, และ premium models ได้ฟรี
No other Thai software house has this structural advantage — $200K USD/year ≈ ฿7,000,000/year in AWS credit to run Claude Opus, Sonnet, and premium models at no cost.
The Numbers
- $100,000 USD/year — Go Digit as AWS Advanced Tier Services Partner
- $100,000 USD/year — Next Brain as separate AWS Partner entity
- Total: $200,000 USD/year ≈ ฿7,000,000/year in AWS credit
- Usable on AWS Bedrock — hosts Claude, Llama, Titan, Cohere, Mistral models
- Renewable annually (subject to partner program status)
What This Means Operationally
We can run Claude Opus, Claude Sonnet, and other premium models for:
- Prototype development during AIRA-X Diagnostics (Day 1-3 builds, not wireframes)
- Engagement delivery (Sprint, Custom Dev, Implementation)
- Internal R&D — testing new AI patterns without burning cash
- Sales POCs — rapid demos without client paying POC fees
- Next Brain product development — training, eval, continuous improvement
Structural Advantage
No other Thai software house has this. Not G-Able, MFEC, Bluebik, or any indie AI shop. Dual-entity structure (Go Digit + Next Brain as separate partners) doubles it.
What This Translates To
| Metric | Competitor (no credit) | Go Digit (w/ credit) |
|---|---|---|
| Claude Opus prototype for 1 day | ~฿15-30K tokens | ฿0 (covered) |
| Weekly R&D experiment budget | Cash-flow limited | Effectively unlimited |
| POC quality during sales | Low-end models | Opus-tier |
| AI code review | Too expensive to automate | Runs on every commit |
| Prompt engineering iteration | 10-20/week | 200-500/week |
| Spec / design AI review | Selective | Standard practice |
→ 5-10x quality/speed advantage on anything touching premium models — paid for by AWS, not us or clients.
🎯 How We Use It (Internal Playbook)
During Diagnostic (Days 1-3)
- Run Claude Opus against client's actual data (anonymized) during Day 2 analysis
- Produce working prototypes on Day 2-3 as part of readout — not mockups, actual running AI
- Why our Diagnostic feels "unusually deep" vs other 3-day assessments
During Sprint / Custom Dev
- Every prompt gets AI code review before commit
- Claude Opus used for architecture review, not just implementation
- Multiple model providers tested per use case (not locked to cheapest tier)
During AI Build & Run Retainer
- Monthly "model upgrade audit" — test newest models against client's existing prompts
- Cost optimization reports leverage our benchmark data from the credit pool
- POCs per quarter use premium models — competitors can't match quality at our price
🗣️ How to Talk About This
✅ Internal framing (team meetings, planning)
- "We have premium model access as a structural advantage — use it aggressively."
- "Don't cheap out on models during R&D — the credit exists for this."
- "This is why we can deliver Sprint in 6-8 weeks when competitors quote 12."
⚠️ External framing (client conversations)
- "We build with premium Claude models by default — we've optimized our stack for it."
- "Our development velocity is higher because we use best-in-class models during iteration."
- "Anthropic / AWS Bedrock are core to our build methodology."
❌ Never say
- ห้ามพูดว่า: "เรามี AWS credit ฟรี $200K"
- "AWS เป็นคนจ่ายค่า development"
- "เราได้ Claude tokens ฟรีจาก AWS"
- หรืออะไรก็ตามที่ทำให้ดูเหมือนว่า advantage ของเราเป็นของ supplier ไม่ใช่ตัวเราเอง
- "We have $200K of free AWS credits"
- "AWS pays for our development"
- "We get free Claude tokens from AWS"
- Anything implying our advantage is supplier-subsidized
Why: Clients who learn you're running on vendor credits will ask "why should I pay full price?" — and partner terms forbid disclosure anyway.
🛡️ Protection Rules
Do NOT Expose In:
- Website copy
- Sales decks (public)
- Pitch emails
- LinkedIn posts
- Press releases
- Case studies
- Investor pitches (written)
- RFP responses (mention tier only, not credit)
Do Reference In:
- Internal team planning
- Internal sales playbooks (this doc)
- Internal capability discussions with Director/CEO
- Acquirer / strategic investor verbal discussions
- Internal hiring pitches ("why you want to work here")
🧠 The Bigger Pattern — Multiple Structural Advantages
Go Digit has accumulated multiple small structural advantages that together form a moat:
- $200K AWS Bedrock credit INTERNAL
- AWS Advanced Tier Services Partner status PUBLIC
- Cloudflare Partner status PUBLIC
- n8n Authorized Reseller — only in Thailand PUBLIC
- Humming Bird Consulting strategic partnership PUBLIC
- Nomura Consulting partnership INTERNAL UNTIL Q3
- Proprietary AIRA-X framework
- Proprietary Next Brain Platform (silent deploy)
No single competitor in Thailand has more than 2-3 of these. We have all 8.
🔐 Credit Management Policy
Tracking
- Finance team tracks monthly Bedrock spend vs credit balance
- Monthly utilization report — shown in exec meetings only
- Forecast annual burn rate quarterly
Allocation Priority
- Client engagements — always covered (never pay out-of-pocket for Bedrock)
- Next Brain platform development — covered
- Sales POCs — covered (max 2 per prospect)
- Internal R&D — covered within quarterly R&D budget
- Experimental / exploratory — CEO approval above $5K USD/month
Renewal Risk
- AWS partner tier reviewed annually
- Maintain partner requirements: certifications, case studies, customer references, delivery competency metrics
- If tier is downgraded, credit shrinks — treat as operational risk
⚠️ If This Leaks
ถ้าข้อมูล $200K หลุดออกไปนอก:
- ห้ามยืนยันตัวเลข
- ตอบว่า "เราเป็น AWS Advanced Tier Services Partner และ build บน AWS Bedrock — partner terms เป็น confidential"
- แจ้ง CEO ทันที
- ตรวจสอบว่าใครมีสิทธิ์เข้าถึง
- ถ้าลูกค้าถาม — redirect ไปที่ outcomes ไม่ใช่ supplier terms
If $200K credit information appears in external channels (web, social, press):
- Do NOT confirm the number
- Respond with: "We are AWS Advanced Tier Services Partner and build heavily on AWS Bedrock — specific partner terms are confidential"
- Alert CEO immediately
- Audit source: who had access, how did it propagate
- If client-facing: pivot conversation to outcomes, not supplier terms
🎯 Rules of Thumb
- ใช้ credit เต็มที่ข้างใน
- ห้ามพูดตัวเลขข้างนอก
- ถ้าถูกถามเรื่อง AWS ให้ตอบด้วย tier name เท่านั้น ("Advanced Tier Services Partner") — ไม่บอก credit amount
- treat หน้านี้เหมือน classified document
- Revisit annually on partner tier renewal.
- Use the credit aggressively internally. Don't ration.
- Never mention the specific dollar amount anywhere public.
- When asked about AWS, answer with tier name ("Advanced Tier Services Partner") — never credit amount.
- Treat this page as internal confidential — shredder-grade if printed.
- Revisit annually on partner tier renewal.
Final note: This credit is one lever. Your talent, framework, platform IP, and partner ecosystem are the real moat. Treat this as one piece of a multi-layer advantage — not the whole thing.
Source: root/11_internal_competitive_edge.md · CONFIDENTIAL — INTERNAL ONLY