Top 10 AI Prompts and Use Cases and in the Financial Services Industry in Puerto Rico

Puerto Rico's financial services scene is entering a practical AI phase: V2A Consulting's 2024 survey shows 84% of local organizations (and 94% of multinationals on the island) have applied AI in at least one business function, with marketing and service operations leading the way and banks modernizing lending and portfolio workflows.

Adoption momentum is real, but experts warn a widening digital divide, strained grid and spotty rural connectivity could throttle real‑time use cases unless infrastructure and skills improve; 59% of respondents cite lack of in‑house expertise as a top barrier.

That's why targeted upskilling and industry-ready programs matter - training like the AI Essentials for Work bootcamp can help Puerto Rico teams move pilots into production and capture value from fraud detection, underwriting, and customer experience automation (see local reporting on governance and infrastructure).

Program	AI Essentials for Work
Length	15 Weeks
Courses	AI at Work: Foundations; Writing AI Prompts; Job Based Practical AI Skills
Cost (early bird)	$3,582
Registration	Register for the Nucamp AI Essentials for Work bootcamp

“A significant 84% of local organizations report having applied AI in at least one business function. More importantly, results suggest that AI is starting to deliver value to Puerto Rican organizations.”

Selection of prompts and use cases was driven by three pragmatic lenses: local demand and readiness (V2A Consulting's 2024 AI survey showing 84% local adoption and function-specific uptake in marketing and service operations), global signal and technology trends from the State of AI Report 2025 - global AI trends, and hands‑on deployment guidance from practical playbooks like the Nucamp Nucamp Back End, SQL, and DevOps with Python syllabus (MLOps best practices guide).

Priorities favored prompts that map to existing adoption patterns (chatbots and document summarization for service and lending), reduce measurable work (speeding loan‑file review that can shave hours from underwriting), and respect local constraints (skills gaps and concerns about privacy, accuracy and IP).

Each candidate prompt was scored for immediate ROI, regulatory sensitivity, and ease of piloting with existing staff and data - so the final Top 10 emphasizes high-impact, production‑ready use cases Puerto Rican teams can staff and evaluate quickly.

Selection Criterion	Evidence / Source
Local adoption signal	84% of local organizations applied AI (V2A Consulting 2024)
Top adopter functions	Marketing & service operations leading uptake; chatbots common (V2A Consulting 2024)
Investment & readiness	37% limited to basic subscriptions; 42% willing to customize; 18% already invested (V2A Consulting 2024)

“A significant 84% of local organizations report having applied AI in at least one business function. More importantly, results suggest that AI is starting to deliver value to Puerto Rican organizations.”

For Puerto Rico lenders, AI-powered chatbots and document‑summarization tools promise real operational relief - automating document collection and OCR, structuring messy paystubs and bank statements into underwriting-ready data, and giving borrowers 24/7 status updates so loan cycles don't stall; industry guides note that the biggest wins come when models translate unstructured data into formats rules engines can act on, while humans handle edge cases and exceptions (mortgage chatbot use cases and vendors research, and practical origination plays).

Yet the Consumer Financial Protection Bureau cautions that chatbots are effective for routine questions but can create costly harms - “doom loops” where customers cannot reach a human, inaccurate answers, or privacy lapses - so firms on the island should pair co‑pilot bots with clear handoffs, Spanish/LEP support, and strict audit trails to meet local compliance and trust needs; when done right this mix speeds origination and frees staff for complex underwriting, not replace them (CFPB report on chatbots in consumer finance).

Metric	Source / Value
U.S. chatbot engagement (2022)	≈98 million users (~37% of population) - CFPB
Estimated annual cost savings	≈$8 billion (~$0.70 saved per interaction) - CFPB

“Poorly designed chatbots or insufficient human support can lead to widespread harm and significantly undermine trust.”

Puerto Rico lenders can widen credit access and sharpen underwriting by blending traditional bureau files with alternative signals - rental and utility histories, payroll-verified income and real-time cash‑flow - so thin‑file applicants aren't automatically screened out; Plaid's research shows alternative data can meaningfully expand the addressable market (an estimated 19 million additional U.S. adults) and even lift outcomes such as higher credit limits and lower current‑to‑late roll rates, while Experian documents how expanded FCRA‑regulated data and custom models can nearly double approvals for some lenders and score far more consumers (Lift Premium covers ~96% vs ~81% for conventional models).

At the same time, responsible use demands explainable models and careful orchestration: FICO notes alternative inputs add predictive value (roughly 60% of the marginal predictive power in one study) but that machine learning must be translated into explainable scorecards and audited workflows to satisfy regulators and consumers.

Practical orchestration platforms - the types Alloy and Plaid describe - let Puerto Rico teams connect payroll, bank and bill‑payment feeds without heavy engineering, enabling faster decisions, better identity checks and ongoing monitoring.

The net result: smarter, faster underwriting that brings financially viable but previously invisible borrowers into the market - if explainability, privacy and data‑quality controls are built in from day one, not retrofitted.

Metric	Value / Source
Additional U.S. adults reachable with alternative data	≈19M - Plaid
Scorable population (Lift Premium vs traditional)	96% vs 81% - Experian
Predictive uplift from alternative data (example)	~60% of marginal predictive power - FICO
Credit outcomes cited (examples)	32% higher credit limits; 30% lower current‑to‑late roll rates - Plaid

For Puerto Rico financial teams, transaction anomaly detection - the ML-driven practice of flagging outliers, contextual spikes or small

smurfing

chains - is now a practical lever to cut losses and speed investigations: models that combine unsupervised clustering, supervised classifiers and real‑time risk scoring can spotlight everything from a lone $200,000 point anomaly to a coordinated sequence of small transfers that mask laundering, then surface alerts for human analysts to triage.

Deploying anomaly detection in AML means pairing continuous transaction monitoring and scenario‑based rules with adaptive models that reduce false positives, preserve audit trails, and support Suspicious Activity Reports; local firms that tie these signals into case workstreams and information‑sharing networks get faster, more defensible outcomes.

The tradeoffs are familiar - data quality, class imbalance and evolving fraud tactics require ongoing tuning and cross‑institution collaboration - but when implemented as part of a layered program, anomaly detection becomes a real‑time safety net that helps Puerto Rico institutions proactively flag threats and lower losses rather than chase them after the fact (see practical primers on anomaly detection in AML and AML transaction monitoring best practices), and local case studies show fraud detection and risk analytics can tangibly reduce loss exposure across the island (fraud detection and risk analytics in Puerto Rico).

Technique	Primary use / benefit	Source
Unsupervised clustering	Detect unknown or collective anomalies	FinancialCrimeAcademy / FraudFights
Supervised models & risk scoring	Classify known fraud patterns, prioritize alerts	FraudFights / Datavisor
Real‑time monitoring + rules	Immediate alerts, SAR support and reduced dwell time	Sanctions.io / Tookitaki

Puerto Rico lenders should treat adverse‑action notices as both a compliance and customer‑care tool: Regulation B (ECOA) and the FCRA demand clear, timely explanations when credit is denied or terms are changed, and the safe path often is a combined notice that meets both regimes.

Key practical rules to build into workflows include Reg B's timing (generally 30 days after a completed or incomplete application or action on an existing account, and 90 days for a counteroffer that isn't accepted), the content checklist (creditor identity, ECOA antidiscrimination language, a statement of action taken and either specific reasons or the applicant's right to request them), and the heightened disclosures when a credit score is used (numerical score, range, key factors and date).

Regulators expect specificity (up to four principal reasons is recommended) and strong controls for automated systems, record retention (most credit applications must be kept 25 months), and third‑party processes; when alternative data or AI/ML inform decisions, Regulation B allows tailored reasons but firms must still explain the actual factors relied on and keep systems tested and auditable.

For hands‑on guidance, see the CFPB's Circular 2023‑03 on proper use of sample forms and the detailed adverse‑action primer from Consumer Compliance Outlook.

Trigger	Typical Timing / Requirement
Completed credit application adverse action	Notify within 30 days (Reg B)
Incomplete application (notice of incompleteness or action)	Notice within 30 days or follow §1002.9(c) procedures
Counteroffer not accepted	Send AAN within 90 days if applicant does not accept/use offer
File retention for credit applications	Retain documents for 25 months

“The requirement that creditors give reasons for adverse action is …. a strong and necessary adjunct to the antidiscrimination purpose of the legislation, for only if creditors know they must explain their decisions will they effectively be discouraged from discriminatory practices.”

Document analysis and due diligence are fertile ground for Puerto Rico lenders ready to turn paper chaos into predictable throughput: AI-driven closers and intelligent document processing can scan a 100‑page closing package, verify homeowner insurance dates, cross‑check names and amounts, and flag a missing notary in seconds - work that once took 30–60 minutes per file - so teams spend minutes on exceptions, not hours on routine checks; platforms like Areal AI Closer Copilot for automated mortgage closings and end‑to‑end automation playbooks for mortgage document workflows such as mortgage document automation workflows show how OCR, NLP and configurable business rules standardize files for LOS ingestion, build audit trails for compliance, and reduce rework that stalls funding.

Practical deployments - from pre‑close collaboration to automated trailing‑document tracking - also cut error rates and speed time‑to‑close dramatically (real case studies report huge drops in human error and faster processing), making it realistic for Puerto Rico teams to scale without hiring at peak volumes while preserving regulatory controls and bilingual borrower communication for local needs.

Metric	Value / Outcome	Source
Typical manual closing review time	30–60 minutes per file	Areal Closer Copilot
Lenders using AI for document processing	≈38%	Docupilot
Case study: error reduction	98% reduction in human intervention errors	MSuite case study
Case study: processing time improvement	10 days → 4 days total processing	MSuite case study
Automation speed claim	Close loans up to 90% faster; save ~10 hours/week on tasks	Addy AI / Addy.so

“Docupilot has been an invaluable tool for streamlining document automation at TLMoto. The platform is intuitive, efficient, and packed with features that make generating documents effortless for any user.” - João S., Head of HR (testimonial)

Localized forecast models and stress tests are essential for Puerto Rico finance teams because island‑specific signals - abundant deposits, unusually strong bank profitability, and the looming phase‑out of federal relief - drive materially different outcomes than mainland models.

V2A Consulting's Q4 2024 banking report shows a striking Pre‑Tax ROE of 20.81% in 2024 and a non‑performing‑loans ratio near 1.75%, indicating banks enter scenarios from a position of capital strength, yet pockets of consumer stress (credit‑card and auto delinquencies) counsel caution; at the same time the Oversight Board's symposium highlights divergent macro paths (Moody's +0.4% FY25 vs the Planning Board's +1.2% and a -0.8% Oversight Board estimate), so prudent scenario planning should combine baseline, funding‑shock and climate/hurricane scenarios to stress liquidity, underwriting and capital returns.

Practical localized models therefore stitch together V2A‑style bank metrics, macro scenarios from the forecasting symposium, and regional indicators from Estudios Técnicos to test whether profitability and capital buffers hold under sudden federal funding losses or a tourism slowdown - so stress tests become planning tools that flag when to conserve capital, tighten underwriting, or tap contingency liquidity.

Metric	Value / Source
Pre‑Tax ROE (2024)	20.81% - V2A Consulting Puerto Rico Banking Report Q4 2024
Non‑Performing Loans Ratio (2024)	1.75% - V2A Consulting Puerto Rico Banking Report Q4 2024
Moody's / Planning Board / Oversight forecasts	Moody's +0.4% FY2025; Planning Board +1.2% FY2025; Oversight Board −0.8% - Puerto Rico Economic Forecasting Symposium (Oversight Board)

“The financial stability index for Puerto Rico's banking industry has recovered after declining from 0.64 in the fourth quarter of 2019 to 0.54 in the first quarter of 2020.” - Estudios Técnicos

Automated reporting is the practical glue that turns island data into investor-ready stories: Puerto Rico finance teams can use platforms that pull from ledgers, loan systems and CRMs to publish scheduled KPI summaries, on-demand dashboards and drill‑down board decks so answers to tough investor questions arrive in minutes instead of days.

Tools such as Domo automated reporting tools bring real‑time visuals, alerts and embedded analytics, while specialist investor tools like Drivetrain investor reporting software automate board packs, scenario analysis and drill‑downs for live Q&A for funds and portfolio managers, programs like Standard Metrics investor metrics platform centralize, auditable metrics and have been shown to slash the time spent collecting and structuring data - turning lengthy monthly close rituals into repeatable, branded report packages that maintain governance and make investor communication a competitive advantage.

The payoff on the island is tangible: fewer spreadsheet bottlenecks, faster responses to LP due diligence, and a crisp single source of truth that helps teams focus on interpretation, not ingestion - imagine handing an investor a live dashboard that answers follow‑up questions on the spot, not a stack of PDFs days later.

Metric / Outcome	Source
Real‑time dashboards, alerts & embedded analytics	Domo
Interactive, drill‑down investor reports & scenario analysis	Drivetrain
95% reduction in time collecting and managing data; 5 hours saved monthly	Standard Metrics

“Standard Metrics saves me a ton of headspace by serving two critical functions; first it puts all my critical data in one dashboard... second, it streamlines my investor relations.” - Nathan Mintz, CEO and Co‑founder

Segmentation and personalization turn scattershot marketing into precise, trust-building engagement for Puerto Rico financial firms: by moving beyond basic demographics to behavioral, technographic and needs‑based segments and layering real‑time signals via a Customer Data Platform, banks and credit unions can deliver offers that matter when they matter - research shows personalization boosts adoption, retention and satisfaction (Appcues research on benefits of personalization in digital marketing) and Deloitte found 80% of consumers prefer personalized experiences and often spend substantially more as a result.

AI‑driven segments uncover hidden micro‑markets (thin‑file borrowers, payroll‑timed savers, or tech‑light customers), enabling targeted offers that expand inclusion while reducing acquisition cost; Tredence urges a shift to intelligent, real‑time personalization for measurable impact.

The practical caveat for Puerto Rico: prioritize first‑party data, consent and clear governance so personalized outreach is helpful not intrusive - think of a timely, bilingual savings nudge synced to a customer's pay cycle rather than a generic email - and pair testing, privacy and CDP hygiene to turn segmentation into scalable financial inclusion (Deloitte personalization strategy playbook for retail and financial services, Puerto Rico financial services AI use cases and cost-saving examples).

For Puerto Rico investors and corporate deal teams, AI is quickly turning cumbersome M&A checklists into high‑velocity, repeatable workflows: tools that auto‑organize virtual data rooms, extract contract clauses, flag tax and indemnity risks, and feed structured inputs into valuation and scenario models let teams spot deal breakers in hours instead of weeks - saving fees and sharpening negotiation leverage.

Practical playbooks show AI excels at mining public filings and VDR content for anomalies and missing documents, and platforms that combine OCR, NLP and model‑led risk scoring can surface change‑of‑control clauses or undisclosed liabilities for human review (see EY overview of AI in due diligence and V7 guide to AI-powered document extraction).

Generative and predictive layers also help simulate purchase‑price outcomes and post‑merger integration scenarios, but local firms should pair automation with strict vendor due‑diligence, data‑privacy controls and lawyer oversight so AI findings become decision‑grade intelligence rather than unsupported assertions (Deloitte guidance on blending Generative AI with human judgment).

“Just as the tech function itself has moved out of the back room to become a strategic centerpiece for many companies, the most effective way to evaluate a target's technology capability has had to evolve as well.” - Bain & Company

Tax teams on the island can turn a perennial headache - manual reconciliations, siloed ledgers and fast‑changing rules - into a competitive advantage by automating core tax, audit and internal‑controls workflows: OCR and AI extract and verify source documents, APIs centralize GLs for “calculation‑first” roll forwards, and agentic automation tracks deadlines and exceptions so humans focus on judgment and advisory work rather than repetitive checks.

Practical playbooks show these tools cut error rates, speed lookback reviews dramatically (EY reports AI‑assisted reviews can be thousands of times faster than manual checks) and free capacity for higher‑value planning, while Thomson Reuters highlights that 79% of professionals expect high impact from AI and many firms already see ROI and measurable time savings - so Puerto Rico finance teams can safeguard compliance (and prepare for changes like expiring TCJA provisions) without ballooning headcount.

Local teams should pair automation with strong audit trails, change controls and upskilling to avoid governance gaps; in one bank deployment robots generated more than a dozen compliance reports in seconds and saved nearly 10,000 staff hours annually, a vivid example of how automation shifts effort from filing to foresight (see practical guidance from Thomson Reuters, EY and the 3Pillar/UiPath case study).

Metric	Value / Source
Professionals expecting high/transformational AI impact	79% - Thomson Reuters
Firms reporting ROI from AI initiatives	54% - Thomson Reuters
Efficiency: weekly hours saved (example)	~5 hours/week; ~$24,000 value per professional/yr - Thomson Reuters
AI‑assisted lookback review speed	Up to ~3,600× faster than human review - EY
Case outcome: hours saved (bank)	~10,000 hours saved annually - 3Pillar / UiPath case study

Puerto Rico financial teams that want to move from pilots to production should treat AI as a program, not a point solution: lean on the three foundations - clear AI governance, strong model risk management and centralized standards - outlined by RSM to reduce regulatory and operational friction, prioritize quick-win pilots (fraud detection, document automation, personalization) that map to local needs, and close the talent gap highlighted by V2A's finding that 59% of organizations lack in‑house expertise; see the V2A Consulting 2024 report for the island's adoption signal and barriers.

Practical next steps include launching scoped proofs‑of‑value tied to measurable KPIs, building explainability and audit trails into models, and investing in targeted upskilling so staff can run and govern systems (training like the AI Essentials for Work bootcamp makes this realistic).

The payoff is concrete - AI can turn a 100‑page closing package into an exception‑only checklist - and with governance and skills in place Puerto Rico firms can capture efficiency, inclusion and risk‑reduction gains fast; read RSM's primer on the three foundations for implementation details.

“A significant 84% of local organizations report having applied AI in at least one business function. More importantly, results suggest that AI is starting to deliver value to Puerto Rican organizations.”