Steering Committee Response Analysis

Customer prospecting triage for AIP sales sprint

✓

Disposition Recommendation

Primary Use

Expert Network Intake

Secondary Use

Sales Language Extraction

Priority Level

Low for Direct Outreach

1 Add enterprise surprise-blocker respondents to Q1 ethnographic interview pipeline
2 Use Chubb/Mastercard multi-respondent signal for account-based outreach to their leadership
3 Extract swipe file quotes into sales enablement materials for CTO conversations
4 Consider parking further analysis unless conference relationship warrants deeper investment

Total Respondents

Conference participants

Billion+ Enterprises

54% of respondents

Strong ICP Fit

Target sector matches

Expert Candidates

Substantive verbatims

◆ Key Findings

ICP Mismatch

Respondents are practitioners (Staff Engineers, Principal Engineers, Directors) rather than budget holders (CTOs, CIOs, VPs of Innovation). Strong influencer pool, weak direct sales targets.

Sector Concentration

Financial services dominates the strong-fit pool: Scotiabank, TD, Commonwealth Bank, Manulife, Prudential, Wealthsimple, Chubb, Mastercard.

Multi-Respondent Signal

Chubb (3 respondents) and Mastercard (2 respondents) indicate organizational depth. Consider account-based outreach to their leadership tier.

Problem Alignment

Dominant blocker theme (evaluation/measurement) maps directly to Q1 "Measuring GenAI Value" study. These practitioners are living the problem AIP researches.

⚠ Competitor presence: Bain & Company and Capgemini respondents flagged in adjacent-fit tier. Treat as competitive intelligence, not prospects.

■ ICP Adjacency Analysis

Strong Fit — Billion+ in Target Sectors 26 respondents

Company	Subsector	Title	Stage
C Chubb3x Insurance	Insurance	Director AVP, Sr GenAI Eng, Sr AI Eng	Production
M Mastercard2x Payments	Payments	Lead Data Scientist, Principal AI Eng	Pilot
S Scotiabank Banking	Banking	Senior Manager, Data Strategy	Pilot
T TD / Layer 6 AI Banking	Banking	Staff ML Scientist	Production
C Commonwealth Bank Banking	Banking	Distinguished AI Scientist	Production
M Manulife Insurance	Insurance	AVP AI	Production
P Prudential Insurance	Insurance	VP, Finance Expense Product Owner	Pilot
G Google Big Tech	Big Tech	Software, ML Lead	Production
M Microsoft Big Tech	Big Tech	Principal Data Scientist	Production
M Meta Big Tech	Big Tech	Sr. Machine Learning Engineer	Production
N Netflix Streaming	Consumer Tech	Sr. Engineer	Production
C Cardinal Health Healthcare	Healthcare Distribution	Data Analytics Manager	Production

Adjacent Fit — Billion+ Non-Target Sectors 6 respondents

Company	Sector	Title	Notes
Bain & Company	Consulting	Expert Senior Director	Competitor
Capgemini	IT Consulting	Chief Architect	Competitor
Fujitsu North America	IT Services	Leader, Emerging Data Tech	Potential referral
Gap	Retail	Sr ML Manager	Potential referral
Hydro One	Utilities	Director Data & AI	Potential referral
Sky	Media/Telecom	Head of Data Platform	Potential referral

★ Expert Network Candidates

High-engagement respondents for ethnographic interviews or Q1 polling. These practitioners provided substantive "surprise blocker" responses.

Enterprise — Healthcare

Internal security review and data classification. Technically everything worked, but it took months to get approval on which internal documents could legally enter prompts, how logs were stored, and whether outputs could be retained — far longer than building the system itself.

Cardinal Health

Data Analytics Manager

Enterprise — IT Services

We encountered a significant regulatory hurdle: the EU AI Act classifies certain emotion recognition systems specifically those used in workplaces as 'prohibited.' This created a compliance blocker for our European operations that does not currently exist in the North American market.

Fujitsu North America

Leader, Emerging Data Technologies

Enterprise — HR Tech

Integrating LLMs or agent-based workflows into older enterprise systems (especially those tightly coupled to databases containing PII) requires far more architectural rethinking and refactoring than initially anticipated. It's not just an API swap.

Dayforce HCM

Senior ML/AI Developer

Enterprise — Payments

Defining, developing KPIs for GenAI models in prod.

Mastercard

Lead Data Scientist

Enterprise — Enterprise Tech

Measuring Agent performance. Convincing users that ML systems do not achieve 100% accuracy and that stochastic processes do fail sometimes, hence need HITL.

Cisco Systems

Principal ML Engineer

Other — Health AI

I was surprised how often RAG would silently fail to provide relevant context and that RAG performance monitoring is very crucial with evolving products/knowledge base.

Amplifai Health

Co-founder & CTO

❖ Language Swipe File

Enterprise practitioner language for CTO/CIO conversations. Use these quotes to demonstrate understanding of production challenges.

Evaluation & Measurement

Evaluation complexity: GenAI/LLMs are non-deterministic, which makes it difficult to build reliable evaluation frameworks. Traditional testing or ML metrics do not translate cleanly, which slows down trust, governance, and sign-off for production use.

Scotiabank

Senior Manager, Data Strategy

Undefined evaluation — Teams can't agree on what 'good' means or tie model output to a business KPI.

Google

Software, ML Lead

Evaluation is hard — Unlike traditional ML where you have clear metrics, evaluating whether an LLM response is 'good' often requires human review or building complex eval frameworks.

AMD

Principal Staff

Reliability & Trust

Hallucinations & reliability. For example code review agent gives incorrect response—it breaks the trust of the user getting the feedback.

Gap

Sr Machine Learning Manager

Reliability/consistency — Models still produce unpredictable outputs that break production workflows. A 95% success rate sounds good until you're handling errors 5% of the time.

AMD

Principal Staff

Organizational & Leadership

Leadership expectations: Senior stakeholders often treat AI as a universal solution rather than a tool with constraints. This leads to misaligned goals, unrealistic timelines, and pressure to 'ship something' before foundations are ready.

Scotiabank

Senior Manager, Data Strategy

Systems thinking is not uniformly present across the teams involved from product to data science to engineering to ops to risk management etc. They don't speak the same language hence all the things that fall through cracks.

Commonwealth Bank of Australia

Distinguished AI Scientist