# SCOPE — NSights

Working scoping doc. Drives MVP decisions before build kickoff. Three sections: full feature backlog, outstanding decisions for Kristian, proposed MVP scope.

---

## 1. Feature backlog

Priority key: **M** = MVP1, **M+** = MVP2/3, **L** = later, **?** = needs decision.

### A. Foundation

| # | Feature | Pri | Notes |
|---|---|---|---|
| A1 | Multi-tenant auth (firm / advisor / admin roles) | M | CIP needs firm-admin role |
| A2 | Firm branding (logo, colours, fonts) | M | Every output is firm-branded |
| A3 | Firm compliance disclosures | M | Per-firm + per-consultant variants |
| A4 | Per-advisor user management | M | |
| A5 | Settings | M | |
| A6 | Demo/trial mode | M+ | "Demo User" badge in mocks |

### B. Data layer

| # | Feature | Pri | Notes |
|---|---|---|---|
| B1 | Monthly fund-price ingestion | M | 90% of analytics derive from this |
| B2 | Direct life-co feeds (Plan A) | ? | Pending ops-to-ops |
| B3 | Manual upload fallback | M | Fund Data Upload nav item exists |
| B4 | Fund metadata (name, strategy, asset class) | M | |
| B5 | ESMA SRRI tagging | M | Charts overlay ESMA bands |
| B6 | SFDR classification | M+ | Mentioned, not in sample |
| B7 | Asset allocation (quarterly) | M | Public sources |
| B8 | Tactical asset allocation calls | M+ | Per-provider, low-freq |
| B9 | Market index data (S&P etc.) | M | Bloomberg + FE today |
| B10 | Historic stress scenarios (one-time) | M+ | 2022 sell-off, COVID, etc. |
| B11 | Wealth-manager funds (Davy etc.) | M+ | Beyond 6 life cos |

### C. Investment Agent ("Warren") — Research & Comparisons

| # | Feature | Pri | Notes |
|---|---|---|---|
| C1 | Fund selection grid with provider tabs | M | |
| C2 | Selected-funds tray + comparison view | M | |
| C3 | Cumulative performance line chart | M | Sample slide type |
| C4 | Annualised performance bars (multi-horizon) | M | |
| C5 | Rolling 2-year performance | M | |
| C6 | Yearly performance bars | M | |
| C7 | Volatility (5Y) computation | M | Confirmed only risk metric |
| C8 | Vol-vs-return scatter + ESMA bands | M | |
| C9 | Peer-group benchmarking | M | Regulatory requirement |
| C10 | Qualitative "why" written commentary | M | "Explain why fund is performing way it is" |
| C11 | Custom charts builder | M+ | Tab visible in mocks |
| C12 | Auto-disclosure compilation | M | |

### D. Portfolio Builder

| # | Feature | Pri | Notes |
|---|---|---|---|
| D1 | Drag-drop portfolio construction | M+ | |
| D2 | Strategy-agnostic engine | M+ | Passive/active/equal-weight/core-sat |
| D3 | Written + quantitative rationale | M+ | "Simple Rationale" modal |
| D4 | Key portfolio metrics panel | M+ | Vol, ESMA, returns |
| D5 | CIP — firm-managed portfolios | M+ | Dashboard table |
| D6 | Investment-committee controls | M+ | Firm-admin only |
| D7 | Rationale propagation to advisors | M+ | Changes flow downstream |
| D8 | "Add to Preview" → deck integration | M+ | |

### E. Planning Agent — Financial Planning

| # | Feature | Pri | Notes |
|---|---|---|---|
| E1 | Goals-based reasoning workflow | L | |
| E2 | Outcome / scenario modelling | L | |
| E3 | Concepts library (visual explainers) | M+ | Time in Market, Mag 7, drawdowns — lifted from Kaben deck |
| E4 | Forecast graphing tool | L | |

### F. Client Support / Compliance Agent — Client Portal

| # | Feature | Pri | Notes |
|---|---|---|---|
| F1 | CRM integration framework | M+ | |
| F2 | First CRM connector | M+ | Which one TBD |
| F3 | AUM Distribution by Provider chart | M+ | |
| F4 | Client Performance Distribution | M+ | |
| F5 | Per-client suitability button | M+ | |
| F6 | Per-client performance button | M+ | |
| F7 | Auto-generated suitability reports | M+ | |
| F8 | Auto-generated performance/review reports | M+ | |
| F9 | Continuous monitoring (background) | L | Cron + intelligent prompts |
| F10 | Performance alert flagging | M+ | |

### G. Marketing Agent — Insights & Campaigns

| # | Feature | Pri | Notes |
|---|---|---|---|
| G1 | Campaign library shell | L | |
| G2 | Campaign: Cash deposit → invest | L | |
| G3 | Campaign: IORPS2 | L | |
| G4 | Campaign: Corporate investments | L | |
| G5 | Campaign: UK pension purchases | L | |
| G6 | Campaign: ARF transfers | L | |
| G7 | Campaign: Pre→post retirement | L | |
| G8 | Campaign: Passive→active | L | |
| G9 | Campaign personalisation | L | |
| G10 | Recent News feed | M+ | Manual curation initially |
| G11 | Recent Presentations history | M | Already in mocks |

### H. Deck Builder / Output Engine

| # | Feature | Pri | Notes |
|---|---|---|---|
| H1 | Persistent right-rail PPT Preview | M | Critical UX pattern |
| H2 | Cover slide template (firm-branded) | M | |
| H3 | "Add Content Slide" composition | M | |
| H4 | End/disclaimer slide | M | |
| H5 | PPTX export | M | Pixel-faithful |
| H6 | PDF export | M+ | |
| H7 | Save deck (drafts) | M | |
| H8 | Share deck (link/email) | M+ | |
| H9 | Server-side slide rendering engine | M | python-pptx / Aspose / HTML→PPTX |

### I. Governance / IP layer

| # | Feature | Pri | Notes |
|---|---|---|---|
| I1 | Whitelisted data sets only | M | The moat |
| I2 | Ireland CPC built-in rule set | M | |
| I3 | Firm-specific rules | M+ | |
| I4 | Per-compliance-consultant disclosure variants | M+ | |
| I5 | Audit log of agent outputs | M+ | Regulatory defensibility |

### J. Future / nice-to-haves

| # | Feature | Pri | Notes |
|---|---|---|---|
| J1 | Application/transfer form storage | L | |
| J2 | In-platform form completion | L | Partnership lever |
| J3 | Multi-region (Canada, SA) | L | 18–24mo horizon |
| J4 | Full white-label | L | |
| J5 | D2C product line | L | |

---

## 2. Outstanding from Kristian

30 decisions. Grouped.

### Strategic & briefing

1. Brief format: investor 1-pager / full product brief / both?
2. Side-note cashflow opportunity (custom reporting for non-ICP firms): A fold-in / B separate line / C park?
3. CTO-advisor questions (agents / IP / first-mover / software meltdown): in brief, or separate response?
4. "Software meltdown" — which incident does he mean?

### Data sourcing

5. Status & timeline of direct life-co feed conversations
6. Fallback path if life cos slip — license FE/Fund Focus, scrape, or hybrid?
7. Initial fund universe — all 6 life cos + Davy, or narrower for MVP?
8. Ops-to-ops meetings — solo or with CTO present?

### Product scope

9. "Warren" — one agent or four (one per Agent)?
10. Risk metrics: volatility only confirmed, but Demo 1 mentions "drawdowns, risk metrics". Resolve.
11. Concepts library: NSights team curates, or AI generates?
12. Compliance disclosures: per firm + per consultant + versioned?
13. Branding scope: logo+colours only, or full theme system?
14. Geographic MVP: Ireland-only, or multi-region day one?

### Integrations

15. CRM priority: which first — Tishkint / BISPlatform / Online Application / Money Advice / Bestadmin?
16. Bestadmin in mocks — confirmed partner or placeholder?
17. CRM commercial terms — anything agreed?

### Commercial

18. Pricing model: flat €3,096/firm/yr or tiered (Solo / Pro / CIP)?
19. Per-seat or per-firm?
20. Onboarding: self-serve or white-glove?
21. Trial period structure?

### Build & technology

22. Mosaic still in conversation, or out?
23. CTO appointment status & timeline?
24. Build budget split (€250k total — what fraction to build vs ops)?
25. Hosting preference (AWS/GCP/Azure, EU-only)?
26. SLM vs general LLM — hard requirement or marketing language?
27. Data residency / GDPR position?

### Regulatory

28. CPC March 2026 amendments — has the rule set been distilled, or do we read source?
29. Compliance-consultant relationships — who validates output disclosures?
30. CRM client-data PII — GDPR strategy (DPIA, processor agreements)?

---

## 3. Proposed MVP scope

**Premise**: 14-month build sequenced to land value early. €250k total budget (€100k EI + €150k seed) is tight; ruthless MVP1 scope is the lever.

### The MVP1 wedge

> An advisor selects 2–6 funds, NSights produces a firm-branded slide deck containing all charts, metrics, peer benchmarks and qualitative commentary needed to satisfy the new CPC research requirement.

That alone justifies €3k/yr and survives a regulator inspection. Everything else is upsell.

### MVP1 — months 0–4 (the wedge)

| Layer | Included |
|---|---|
| Foundation | A1–A5 |
| Data | B1, B3, B4, B5, B7, B9 |
| Investment Agent | C1–C10, C12 — all 4 risk tiers |
| Deck Builder | H1–H5, H7, H9 — PPTX out |
| Governance | I1, I2 |

**Cost estimate**: ~€60k (4 months × 1.5 senior devs).

### MVP2 — months 4–8 (CIP + compliance starter)

| Layer | Included |
|---|---|
| Portfolio Builder | D1–D5, D8 |
| Compliance | F1, F2 (one CRM), F3–F8 — basic reports |
| Concepts | E3 — 6–10 curated visuals from existing Kaben deck |
| Output | H6 — PDF export |

**Cost estimate**: ~€80k (4 months × 2 devs).

### MVP3 — months 8–14 (depth)

- Marketing Agent + 5 named campaigns (G2–G6)
- Continuous monitoring (F9, F10) — background jobs
- Recent News feed (G10)
- 2nd CRM connector
- Investment-committee controls (D6, D7)
- Custom chart builder (C11)
- Audit log (I5)

**Cost estimate**: ~€100k (6 months × 2 devs).

**Total**: ~€240k over 14 months. Fits €250k with ~4% margin. Tight.

### Explicitly out

- Planning Agent / scenario modelling (E1, E2, E4)
- Application forms (J1, J2)
- Multi-region (J3)
- Full white-label (J4)
- D2C (J5)

### Why this shape

1. MVP1 alone is sellable — CPC is the wedge, everything else is depth.
2. CRM integration is high-effort, low-leverage until first cohort proves value. Push to MVP2.
3. Continuous monitoring is the most expensive feature — defer until customers + data exist.
4. CIP in MVP2 unlocks larger firms → pricing tier upgrade.
5. Marketing Agent gives smallest moat per build hour — defer to MVP3.