All method demos

Tier 2 · Included in base scope

NLP Topic Modeling on Stakeholder Transcripts

BERTopic + LDA models surface latent themes across 129+ transcripts. Every AI-flagged theme is reviewed and validated, refined, or rejected by experts — full audit trail.

Top Surface Themes — Validated by Experts · 129 stakeholder transcripts analyzed

AI-surfaced themes confirm what the interview team expected.

Workforce talent pipeline & skills gaps

talent pipelineskills gaptech workershiring difficulty

47%

of transcripts

Validated

Cost of living & housing affordability

afford to livehousing pricesstay in St. Pete

42%

of transcripts

Validated

Public transit & mobility

SunRunnercross-baycommute time

38%

of transcripts

Validated

Downtown vibrancy & cultural tourism

walkabilityPierMural District

35%

of transcripts

Validated

What this scenario shows

The most prominent topics from BERTopic analysis of 129 stakeholder transcripts align with what the expert team would have identified manually. AI confirms — and quantifies — workforce gaps, cost-of-living concerns, and infrastructure needs as dominant themes. All validated by expert review.

AI/Expert Reconciliation Insight

Validation case: AI confirms what experts expect, providing quantitative anchoring. Useful as 'show your work' material for City Council.

Illustrative dummy data

All figures shown are illustrative and chosen to demonstrate three meaningfully different findings. Final analysis at engagement award is derived from authoritative sources (BLS, ACS, Florida DEO, IMPLAN, CoStar) with full methodology documentation.

Back to all method demos