AI Strategy Package

Heritage Community Credit Union

Prepared by Seiche Advisors
Engagement: Journey · August 5–16, 2026
Lead consultant: Maria Santos
Classification: Proprietary & Confidential

Table of contents

1. Executive summary
2. Institution profile
3. Current-state assessment
4. Maturity scoring
5. Opportunity map
6. 12-month roadmap
7. Pilot specifications
8. AI governance framework
9. Policy recommendations
10. Vendor landscape
11. Budget & ROI model
12. Change management
13. Regulatory alignment
14. Board actions requested
15. Appendices

1. Executive summary

Heritage Community Credit Union ($1.8B, 42,300 members, Eugene OR) is positioned to become an AI-ready community institution within 12 months without hiring data scientists or replacing core systems. The credit union has foundational technology (Symitar, Alkami, Abrigo BAM+, M365) and executive alignment; the primary gap is governance and sequenced execution.

Key findings

• AI is already in use — fraud monitoring, digital FAQ, vendor-embedded models — but undocumented.
• Governance lags peers — no AI policy, no inventory, no designated AI governance lead until August 2026.
• BSA/AML capacity is the binding constraint — highest-ROI first investment.
• Shadow GenAI risk is real — 27% of surveyed staff reported personal AI tool use with work content.

Recommended strategy

Governance first, two governed pilots, scale on evidence.

Year 1 total (midpoint): $225K · Expected ROI: 2.5–3.5x within 18 months.

2. Institution profile

Strategic context

HCCU competes with three regional banks and two larger credit unions in its footprint. Digital engagement scores trail top-quartile CUs by 8 points (Alkami analytics, Q1 2026). Management's 2026–2028 strategic plan cites operational efficiency and member experience — AI supports both if governed.

3. Current-state assessment

3.1 AI inventory (as of August 2026)

Tier definitions: Per Seiche AI Policy Template — Tier 1 (low), Tier 2 (moderate), Tier 3 (high/regulated).

3.2 Organizational readiness

3.3 Stakeholder input (Journey interviews)

4. Maturity scoring

Aggregate from Pre-Compass Questionnaire (8 executives, July 2026):

Heat map summary: Governance is the outlier low score; technology slightly ahead due to Symitar/Abrigo embedded ML.

Target by Q2 2027: Governance L3, Strategy L3, Technology L3, Culture L3.

5. Opportunity map

Prioritized by ROI × feasibility × regulatory acceptability for HCCU.

Department opportunity detail

BSA/AML (Priority 1)
1,400 alerts/month; 62% analyst time on false positives. Verafin or Abrigo AI layer — see Pilot Spec #1.

Lending
Commercial packages: 4.2 hrs avg prep. Ocrolus/nCino viable after policy and fair-lending framework — Q2 2027 earliest.

Member service
12% FAQ deflection today. Authenticated AI (Access Softek IVA, interface.ai) requires policy Tier 2 review and member consent model — 2027.

Operations
Check image fraud, deposit hold automation — core-vendor roadmap; monitor Jack Henry releases.

6. 12-month roadmap

Q3 2026 — Foundation ("Get Smart, Get Safe")

Q4 2026 — Deploy pilots

Q1 2027 — Measure & decide

Q2 2027 — Scale (conditional)

Q3'26        Q4'26        Q1'27        Q2'27
[Foundation] [Pilot live] [Evaluate]   [Scale?]
     |            |            |            |
  Policy      BSA POC      Board        Full BSA
  Inventory   Copilot      decision     + Copilot

7. Pilot specifications

Full specifications are standalone documents:

Sequencing rule: Pilot #2 starts only after AI acceptable-use policy effective date.

8. AI governance framework

HCCU's full AI Center of Excellence (COE) framework — organizational model, RACI, intake workflow, KPIs, 90-day launch plan, and board reporting templates — is delivered as a standalone Journey artifact: 10-ai-coe-framework.html.

Summary (see COE framework for detail)

Three-tier review process

9. Policy recommendations

HCCU should adopt the Seiche AI Acceptable Use Policy Template with these institution-specific customizations:

Target board ratification: September 2026 board meeting.

Policy owner: AI Governance Committee · Annual review: August each year.

10. Vendor landscape

Shortlist by use case (community CU–relevant)

BSA/AML

GenAI productivity

Digital / member AI (2027)

Full directory: Seiche AI Vendor Directory (ingested in portal).

11. Budget & ROI model

Year 1 — conservative

Board request top line: $265K includes Year 1 scale prep.

ROI model (18-month)

Benefit assumptions documented in Appendix C — conservative 50% haircut on vendor case studies.

12. Change management

Messaging framework

Lead message: "AI makes your best work easier — it does not replace the Heritage relationship."

Known resistance

• BSA analysts: Fear of blame for AI misses → human-in-the-loop + override culture
• Lending: Fair lending anxiety → no lending AI in Year 1
• IT: Integration burden → Seiche PMO + vendor professional services

13. Regulatory alignment

Examiner narrative (draft):
"HCCU maintains an AI inventory, acceptable-use policy, and governance committee. Tier 3 AI is limited to BSA workflows with mandatory human review. POC validation documentation available on request."

14. Board actions requested

1. Authorize Seiche Journey net fee ($20,000 after Compass credit)
2. Approve Year 1 AI program budget up to $265,000
3. Ratify AI acceptable-use policy at September 2026 meeting
4. Receive quarterly AI progress reports starting Q4 2026

See Board One-Pager (04-board-one-pager.html) for vote-ready resolution language.

15. Appendices

Appendix A — Questionnaire aggregate heat map

              L1    L2    L3    L4
Strategy       |    ●●   ●    |
Governance    ●●    |    |    |
Technology     |    ●●   ●    |
Culture        |    ●●   |    |

(● = executive count at level; mode governance L1)

Appendix B — AI inventory worksheet (blank template)

Appendix C — ROI assumption sources

Appendix D — Document index

Package complete. ~30 pages formatted for board distribution.

© 2026 Seiche Advisors, LLC. All rights reserved. · seicheadvisors.com · Proprietary & Confidential