clawdbot-workspace/proposals/solvr-discovery-questions-v2.md
2026-02-06 23:01:30 -05:00

433 lines
16 KiB
Markdown

# Solvr Discovery Questions — V2
## Expanded Scope Pre-Engagement Call
*Updated for the full 4-pillar, 47-service implementation. Covers everything from the original plus clinical, financial, and scale questions.*
---
## 1. Pillar Prioritization & Strategy
**We've scoped Growth Accelerator as Phase 1. Does that align with your priorities, or would you rather lead with a different pillar?**
*Why: The whole timeline is built around this. If they want Finance first, we rearrange.*
**Which 3-5 services are the ones your clients ask about MOST?**
*Why: Whatever they name, we make sure those are bulletproof in Phase 1.*
**Are there any services in the catalog you're NOT currently delivering (aspirational) vs. ones you're actively doing today?**
*Why: Huge difference between automating something that exists vs. building something from scratch. Affects complexity.*
**How many hospice agency clients are you currently serving?**
*Why: Scopes the multi-tenant requirements. 3 clients = simple. 30 = needs serious data isolation.*
**What's your target client count in 12 months?**
*Why: Architecture decisions change if they want to 5x their client base.*
---
## 2. User Interface & Access
**How do you and your team prefer to interact with agents day-to-day?**
- Telegram?
- iMessage?
- Slack?
- Email?
- Web dashboard?
*Why: Different interfaces = different infrastructure. iMessage needs Mac hosting. Telegram is simplest.*
**Will your team members interact with specific agents, or do they need access to all of them?**
*Why: 8 agents is a lot. Most people should probably interact with 1-2 that are relevant to their role.*
**For client-facing tools (Website Chatbot, Marketing Bot, HR Hotline) — what interface do your clients expect?**
- Embedded on their website?
- Standalone portal?
- SMS/text-based?
- All of the above?
*Why: Each interface is a different build. Web embed is simplest. SMS routing through GHL is also clean.*
**Do you need mobile access, desktop, or both?**
*Why: Mobile-first means Telegram/iMessage. Desktop opens up more options.*
---
## 3. Infrastructure & Hosting
**Do you have a preference for where the platform runs?**
- Your own servers (we deploy to your infrastructure)
- Cloud instance we manage for you
- No preference — recommend what's best
*Why: 8 agents + 16 MCP servers = more infrastructure than the original 3-agent setup. Needs proper scoping.*
**Are there any compliance or data residency requirements?**
- SOC 2?
- HIPAA? (almost certainly yes, given clinical services)
- Data must stay in US?
*Why: If HIPAA applies (and it likely does for Clinical Eligibility AI), hosting choices narrow significantly.*
**How do you currently handle API credentials and secrets?**
- 1Password / secrets manager?
- Shared doc?
- IT team manages?
*Why: With 16 MCP servers needing credentials, this matters a lot.*
---
## 4. HIPAA & Clinical Compliance (NEW — Critical)
**Does Solvr currently have a HIPAA compliance program in place?**
- BAA templates?
- Designated Privacy Officer?
- Existing risk assessments?
*Why: Clinical Eligibility AI and clinical auditing touch PHI. We need to know the compliance baseline.*
**For Clinical Eligibility AI — what patient data would be ingested?**
- Diagnoses?
- Lab results?
- Physician notes?
- Referral forms?
*Why: Scopes what the AI needs to process and what HIPAA safeguards are required.*
**Do your clients' EMR/EHR systems have API access?**
- Which systems? (Axxess, MatrixCare, Homecare Homebase, Brightree, Netsmart, etc.)
- Read-only or read/write?
- Do they allow third-party integrations?
*Why: If we can pull data via API, automation is clean. If it's manual export/import, that changes the workflow significantly.*
**Are you comfortable with Clinical Eligibility AI being ADVISORY ONLY (human clinician always makes the final call)?**
*Why: Non-negotiable from our side, but want to confirm alignment. No AI should be making clinical eligibility determinations autonomously.*
**Who on your team (or your clients' teams) would serve as the clinical reviewer for AI-generated outputs?**
*Why: The human-in-the-loop needs to be a specific, qualified person. Not "someone on the team."*
**Do you need a BAA (Business Associate Agreement) with OpenClaw for this engagement?**
*Why: If we're touching PHI in any capacity — even building the system — a BAA may be required. Get legal guidance.*
---
## 5. Financial Systems Integration (NEW)
**What accounting/financial system do your clients use?**
- QuickBooks Online?
- QuickBooks Desktop?
- Sage?
- Other?
*Why: QuickBooks Online has a solid API. Desktop is much harder. Sage is a different integration entirely.*
**Is financial data currently centralized, or does each client manage their own books?**
*Why: If Solvr has access to client QuickBooks instances, we can automate. If clients manage independently, we need a data ingestion workflow.*
**For PPD Expense Analysis and CAP Monitoring — where does the source data live today?**
- Spreadsheets?
- Accounting system?
- EMR reports?
- Multiple places?
*Why: "Multiple places" = we need data normalization. Single source = much cleaner.*
**Do you generate financial reports in a specific format today? Can you share examples?**
*Why: Easier to match existing formats than invent new ones. Reduces client confusion.*
**For AP Management — who currently approves payments, and what's the approval threshold?**
*Why: Affects the human-in-the-loop design. $500 approval threshold vs. $5,000 = different workflows.*
---
## 6. GoHighLevel Specifics
**Which GHL features are mission-critical across your services?**
- Contacts & pipelines?
- SMS/email campaigns?
- Workflows & automations?
- Reporting?
- All of the above?
*Why: You have 65+ tools built. This confirms coverage.*
**Are you on GHL Agency or Location level?**
*Why: Agency = more powerful API access. Location = more limited.*
**Do your individual hospice clients have their own GHL sub-accounts, or does everything run under Solvr's account?**
*Why: Multi-tenant GHL setup affects how agents interact with the system. Sub-accounts = cleaner isolation.*
**Do agents need to trigger GHL workflows, or just read/write data?**
*Why: Triggering workflows is more complex. Reading/writing is straightforward.*
---
## 7. Knowledge Base & Content (NEW)
**Do you have existing libraries of:**
- Hospice regulatory reference materials?
- Training curricula / educational content?
- Policy & procedure templates?
- Clinical form templates?
- Best practice guides?
*Why: The Knowledge Base MCP is only as good as the content we feed it. If you have 500 pages of SOPs, great. If we're starting from scratch, that's a different project.*
**How often does hospice regulatory content change?**
*Why: Affects how we design the knowledge base update workflow. Annually = simple. Monthly CMS updates = needs automated monitoring.*
**For the Inservice Suite — do you have existing inservice presentations, or should the AI generate them from scratch?**
*Why: Generating from existing content = higher quality, faster. From scratch = needs more prompt engineering and review cycles.*
**Who's the subject matter expert for clinical/regulatory content?**
*Why: Someone needs to validate what the Knowledge Base says. AI can structure and surface content, but a human expert needs to verify accuracy.*
---
## 8. LinkedIn & Social Media
**For LinkedIn management — are you comfortable with agents drafting posts for human approval, or do you want fully automated posting?**
*Why: Human approval = safe. Fully automated = ban risk.*
**Would you use a dedicated LinkedIn account for automation, or personal accounts (Zack/Allyson)?**
*Why: Dedicated = safer. Personal = riskier but more authentic.*
**Are you aware that LinkedIn actively restricts automation, and there's inherent account risk?**
*Why: CYA. Get this on record.*
**What social platforms beyond LinkedIn matter for your clients?**
- Facebook?
- Instagram?
- TikTok?
- None — it's all LinkedIn?
*Why: Scopes the social media automation build.*
---
## 9. Meeting Intelligence & Coaching
**Where are meeting recordings stored today?**
- Google Drive?
- Local downloads?
- Zoom cloud?
*Why: Scopes the ingestion pipeline for the Whisper MCP.*
**For the Sales Enablement Package (podcasts, coaching, role-play) — is this for Solvr's internal team or delivered to clients?**
*Why: Internal = simpler. Client-facing = needs white-labeling and permissions.*
**Do you need real-time transcription or is post-meeting processing fine?**
*Why: Real-time = much harder. Post-meeting = standard whisper workflow.*
---
## 10. Dashboards & Reporting
**For the Operational Dashboard — what does the ideal dashboard look like?**
- Real-time metrics?
- Daily/weekly snapshots?
- Client-facing vs. internal?
*Why: Real-time dashboards are architecturally different from periodic reports. Need to scope this correctly.*
**Do your clients currently receive reports from you? In what format?**
- PDF?
- Email with data?
- Portal / dashboard access?
- Presentations?
*Why: Affects how the Dashboard/Reporting MCP outputs are formatted and delivered.*
**What are the top 5 KPIs your clients care about most?**
*Why: Dashboard should lead with what matters, not drown in data.*
---
## 11. Talent Solutions Specifics (NEW)
**For the Hiring for Excellence System — do your clients use any ATS (Applicant Tracking System) today?**
- Indeed?
- ZipRecruiter?
- GHL as ATS?
- Manual process?
*Why: If they have an ATS, we integrate. If they use GHL, even better (already built). Manual = we build the system from scratch.*
**For the HR Knowledge Hub — is this based on a standard set of policies, or customized per client?**
*Why: Standard = one knowledge base, many clients. Customized = per-client knowledge base instances.*
**For the HR & Compliance Hotline — what's the expected volume?**
- A few questions a week per client?
- Daily inquiries?
*Why: Affects infrastructure sizing and cost projections.*
**Do your clients have existing employee handbooks we can ingest, or do you create them?**
*Why: Existing handbooks = instant Knowledge Base content. Creating them = a service the AI can also help with.*
---
## 12. Agent Coordination & Autonomy
**With 8 agents, how much autonomy should they have?**
- Fully autonomous for routine tasks?
- Human approval for everything?
- Somewhere in between?
*Why: "Somewhere in between" is the right answer, but we need to define the line.*
**When an agent hands off to another, should that be:**
- Silent (happens in background)?
- Notification to you?
- Approval required?
*Why: 8 agents talking to each other = a lot of handoffs. Silent is simpler but less visibility.*
**Is there a single "source of truth" for client/contact info?**
- GHL is the master?
- QuickBooks for financial?
- Multiple systems?
*Why: With 16 MCP servers touching data, we need clear data authority rules.*
---
## 13. Scale & Multi-Tenant
**When you onboard a new hospice client, what's the process today?**
- How long does it take?
- How much is standardized vs. custom?
- Who does the onboarding?
*Why: One of the biggest ROI wins is automating client onboarding. Need to understand the current pain.*
**Do all clients get the same services, or is it à la carte?**
*Why: All-in = simpler platform config. À la carte = need service toggling per client.*
**What's the maximum number of clients you'd want this platform to support simultaneously?**
*Why: Architecture decisions change at 10 vs. 50 vs. 100+ clients.*
---
## 14. Success Criteria
**What does success look like 90 days after Phase 1 launch?**
*Why: Gets them to define the goal. We deliver to THAT.*
**What's the #1 service you want off your plate?**
*Why: Prioritizes what matters. If we nail this one thing, they're happy.*
**How will you measure ROI?**
- Time saved?
- Clients added without hiring?
- Revenue per employee?
- Client satisfaction scores?
*Why: Anchors expectations to measurable outcomes.*
---
## 15. Timeline & Budget
**Is there a hard deadline or event driving the project?**
*Why: If yes, we know the pressure. If no, we have flexibility.*
**Are you comfortable with the phased approach (Phase 1 live in ~5 weeks, full platform in ~16 weeks)?**
*Why: Some clients want everything at once. That's not how this works — confirm alignment.*
**For pricing — is the phased payment structure workable, or do you prefer a different arrangement?**
*Why: Cash flow matters. Be flexible on structure, firm on total.*
**Do you have budget for the ongoing retainer + infrastructure costs (~$3,500-5,500/month total)?**
*Why: The platform needs ongoing care. Make sure they're not expecting a one-time build with zero maintenance.*
---
## 16. Red Flag Detectors
**Have you worked with AI agents or automation before?**
- If yes: What worked? What didn't?
- If no: What are your expectations?
*Why: Experienced = realistic. First-timers = may need education.*
**Who's the decision-maker for approving deliverables at each phase?**
*Why: With 4 phases, we need clear sign-off authority.*
**Is there anyone else on your team who should be involved in the discovery call?**
*Why: Surfaces hidden stakeholders. Especially important if there's a clinical lead or CFO we should loop in.*
**Are there any services in the catalog you're considering dropping or significantly changing?**
*Why: Don't want to build automation for a service they're about to sunset.*
---
## Questions YOU Should Be Ready For
| Their Question | Your Answer |
|----------------|-------------|
| "Can you guarantee HIPAA compliance?" | "We build with HIPAA safeguards (audit logs, PHI isolation, human-in-the-loop for clinical data). But HIPAA compliance is an organizational commitment — we recommend you consult a healthcare compliance attorney for the overall program. We'll build the technical controls." |
| "Why 8 agents instead of 3?" | "Your service catalog has 47 services across 4 distinct pillars. Three agents would be stretched thin and slower. Eight gives each pillar dedicated intelligence that speaks the right language — clinical, financial, marketing, or HR." |
| "What if LinkedIn bans the account?" | "We mitigate with rate limiting, human approval, and API-first approach. If you'd prefer, we use a dedicated brand account. We'll document the risk and get your written acknowledgment before activation." |
| "$57,500 is a lot." | "It is. But you're building a platform that replaces 3-7 FTEs ($150K-$350K/year in salary). The platform pays for itself within 3-6 months — and then it's just retainer + infrastructure costs going forward. Plus you can 3x your client capacity without 3x the headcount." |
| "Can we start smaller and add on?" | "Absolutely — that's exactly what the phased approach does. Phase 1 gets you live with Growth Accelerator in 5 weeks. You'll be seeing real results before we even start Phase 2. You could technically pause after any phase." |
| "What happens if you get hit by a bus?" | "All code and documentation is yours. Full handoff and training included so your team can maintain it. The MCP architecture is open standard — any developer can work with it." |
| "How do we handle HIPAA for the clinical AI?" | "Clinical Eligibility AI is advisory only — always requires human clinical review. We don't store PHI in AI memory. Everything is audit-logged. We recommend a BAA between our organizations and we'll build to whatever your compliance attorney recommends." |
---
## After the Call
Send a summary email with:
1. Confirmed pillar prioritization
2. Scope inclusions and exclusions
3. Clinical/HIPAA decisions made
4. Financial system integrations confirmed
5. Timeline alignment
6. Next steps and kickoff date
*This protects everyone. If scope creeps later, point to the email.*
---
*Let's build something that changes how hospice consulting works.* 🚀