✏️Prompts

AI Playbook
for Accounting

Tools. Workflows. Controls. Implementation. A practical guide for accountants and finance teams adopting AI responsibly.

How to use this playbook
Read linearly or jump to your role above. All interactive elements save to your browser.

Why AI Matters in Accounting

Real impact metrics and honest limitations. AI is powerful in finance—when designed with controls.

Time Savings
  • 60-80% faster invoice processing (AP teams)
  • 25-40% faster month-end close with AI tools
  • 50%+ reduction in manual reconciliation time
  • 5-10 hrs/week saved on variance explanations
Quality Gains
  • Fewer transposition errors in data entry
  • Consistent application of policies
  • Early detection of anomalies & fraud
  • Better audit trail documentation
Visibility & Control
  • Real-time cash flow forecasting
  • Predictive collection insights
  • Variance drivers identified automatically
  • Exception-based management dashboards
Where AI Fails
  • Complex, non-standard contracts
  • Judgment calls on accounting treatment
  • Fraud detection (needs humans)
  • New / unprecedented transactions
Key principle: AI augments, not replaces
Human oversight required on all financial decisions. AI handles repetitive, pattern-based work.

The Core AI Accounting Stack

Where AI fits across accounting workflows. Seven layers, each with use cases, tools, and risks.

ERP / GL Layer
  • Journal entry suggestions
  • Anomaly flagging in GL
  • Compliance rule checking
Oracle AISAP JouleNetSuite AI
See all tools →
AP Automation
  • Invoice OCR & extraction
  • 3-way match automation
  • Duplicate & fraud detection
BillTipaltiStampli
See all tools →
AR & Collections
  • Payment behavior prediction
  • Collections automation
  • Customer risk scoring
HighRadiusTesorioYayPay
See all tools →
Close & Recs
  • Reconciliation automation
  • Variance explanations (draft)
  • Close checklist orchestration
FloQastBlackLineTrintech
See all tools →
FP&A & Reporting
  • Scenario modeling
  • Budget forecasting
  • Narrative summaries for reports
AnaplanPlanfulPigment
See all tools →
Expense & Spend
  • Policy compliance checking
  • Receipt categorization
  • Vendor analysis
RampBrexNavan
See all tools →
AI Assistants & LLMs
  • Prompt-based analysis & drafting
  • Data review & summarization
  • Policy & memo writing
ChatGPTClaudeCopilot
See all tools →
Risks Across Layers
  • Data privacy & confidentiality
  • Over-reliance without controls
  • Model drift & accuracy decay
  • Audit trail requirements
Architecture tip
AI works best when layered—ERP + AP + AR + Close tools = integrated workflow, not point solutions.

AI for Accounts Payable

Deep Dive

The most mature AI use case in accounting. Invoice-to-cash gets 60-80% faster with proper controls.

Invoice OCR & Extraction
  • What AI does: Reads invoice PDF/image, extracts line items, amounts, vendor, dates
  • Accuracy: 95%+ on standard formats
  • Human review: Always required for payment
3-Way Match Automation
  • What AI does: Matches PO → Receipt → Invoice automatically
  • Flags exceptions: Quantity, price, or date mismatches
  • Reduces: Manual matching time by 70%+
Duplicate Detection
  • What AI does: Identifies duplicate invoices or duplicate line items
  • Prevents: Duplicate payments (high fraud risk)
  • Control: Flag for AP review before approval
Fraud Flagging
  • What AI does: Detects anomalies (unusual vendor, amount, timing)
  • Limitations: Does not prevent fraud—only alerts
  • Must have: Human investigation for flagged items
Approval Routing
  • What AI does: Routes invoices to correct approvers based on rules
  • Accelerates: Approval cycle by 40-50%
  • Maintains: Segregation of duties & audit trail
Vendor Risk Scoring
  • What AI does: Flags new or high-risk vendors
  • Factors: Payment history, concentration, location
  • Compliance: Supports sanctions / OFAC checks

AP Implementation Checklist

Workflow
0 of 10 completed

Pre-Implementation

Post-Implementation

Human-in-the-loop: All invoices over threshold must have manual approval before payment

Audit trail: Log all OCR output, matching decisions, and fraud flags with timestamp & user

Segregation of duties: Approval reviewer ≠ Requester. Payment authorizer ≠ Approver

Data privacy: Mask PII (SSN, bank account numbers) in stored prompts & outputs

Accuracy monitoring: Spot-check 5% of AI-processed invoices monthly; escalate errors to vendor

Reconciliation: Verify all AI-flagged exceptions were investigated & resolved

Fraud response: Define escalation path for suspected duplicate/fraudulent invoices

Top AP vendors
BillTipaltiStampliVic.aiRampAirbaseMineralTreeCoupaBrex

AI for Accounts Receivable & Cash

Deep Dive

Predictive insights + automation for collections, cash forecasting, and customer risk.

Payment Behavior Prediction
  • What AI does: Predicts customer payment likelihood based on history & behavior
  • Improves: Collection timing & resource prioritization
  • Typical accuracy: 75-85% on 30-day prediction window
Collections Automation
  • What AI does: Drafts collection emails, escalation sequences, payment reminders
  • Personalization: Tailors message to payment risk & customer segment
  • Control: Human review before send; tone & legal review required
Customer Risk Scoring
  • What AI does: Assigns risk score based on payment history, industry, size
  • Flags: High-risk customers for tighter credit terms or earlier collection
  • Updates: Dynamically as new payment data comes in
Cash Flow Forecasting
  • What AI does: Predicts weekly/monthly cash inflows using payment patterns
  • Enables: Better working capital planning & liquidity management
  • Accuracy: Improves with more historical data (6+ months)
Dispute Categorization
  • What AI does: Auto-classifies disputes (quality, quantity, pricing, documentation)
  • Routes: To correct resolver (sales, support, finance)
  • Reduces: Manual triage time by 60%+
Deductions & Analysis
  • What AI does: Analyzes deduction reasons & suggests responses
  • Flags: Suspicious or repetitive deductions for investigation
  • Improves: Recovery rate & customer relationship intelligence

AR & Cash Implementation Checklist

Worksheet
0 of 10 completed

Planning Phase

Monitoring & Control

Collections communication: All AI-drafted messages reviewed by senior AR person before send

Escalation protocol: Define thresholds for human intervention (e.g., disputes >10K, legal escalation)

Customer communication: Maintain tone & brand consistency; no aggressive/automated feel

Dispute resolution: Track all disputed items; AI categorization is advisory only

Risk scoring: Review high-risk scores monthly; override if business relationship warrants

Data privacy: Mask PII in AI outputs; restrict forecast data access to finance team only

Model monitoring: Measure forecast accuracy monthly; retrain if accuracy drops >5%

Top AR vendors
HighRadiusTesorioVersapayYayPayGavitiKollenoUpflow

AI for Month-End Close & Reconciliation

Deep Dive

Accelerate close cycles by 25-40%. AI drafts explanations, flags anomalies, orchestrates checklist.

Journal Entry Suggestions
  • What AI does: Suggests accruals, reversals, reclassifications based on patterns
  • Speed: Reduces manual journal entry drafting by 50%+
  • Control: Always human-reviewed & approved before posting
Variance Explanation Drafting
  • What AI does: Drafts narrative explanations for budget vs. actual variances
  • Inputs: KPI table, drivers, one-time items
  • Quality: 70-80% requires human editing; catches 90% of obvious drivers
Anomaly Detection
  • What AI does: Flags unusual GL balances, journal entries, intercompany transactions
  • Detects: Mispostings, duplicates, roundtrip errors
  • False positives: Expect 10-20%; requires human investigation
Reconciliation Automation
  • What AI does: Auto-matches bank transactions to GL, identifies unreconciled items
  • Coverage: 95%+ of standard/recurring reconciliations
  • Edge cases: Manual review required for unusual/multi-month exceptions
Close Checklist Orchestration
  • What AI does: Tracks close task completion, escalates overdue items
  • Visibility: Real-time dashboard of close status by task/owner
  • Speed: Reduces close cycle by 2-5 days typically
Exception Triage
  • What AI does: Categorizes reconciliation exceptions (timing, missing docs, errors)
  • Recommends: Resolution approach & owner for each exception
  • Control: Human signs off on resolution strategy

Close & Reconciliation Control Checklist

Controls
0 of 10 completed

Pre-Close Setup

Close Cycle Execution

Do NOT: Post journal entries without human approval. AI can suggest, never auto-post.

Do NOT: Make accounting judgment calls (capitalization, revenue recognition, fair value adjustments). Humans decide.

Do NOT: Rely on AI-generated variance explanations for board reporting. Always have finance team review & add context.

Do NOT: Override controls for "convenience". If AI can't explain a decision, human must.

Do NOT: Automate away reconciliation reviews. Control owner must visually verify key reconciliations.

Do NOT: Use AI to mask or explain away internal control deficiencies. Fix controls first.

Do NOT: Deduct AI decision time from close budget without adding review/validation time.

Top Close vendors
FloQastBlackLineTrintechNumericAdraDucoReconArt

AI for Reporting & FP&A

Deep Dive

Scenario modeling, budget forecasting, narrative summaries. AI speeds analysis; humans own interpretation.

Scenario Modeling
  • What AI does: Runs sensitivity analyses, models "what-if" scenarios at scale
  • Speed: Tests 100+ scenarios in hours vs. days manually
  • Limitation: Assumes historical patterns; unpredictable events still need human judgment
Budget Forecasting
  • What AI does: Projects revenue, expense, headcount based on trends & drivers
  • Accuracy: 85-95% for recurring items; worse for new/volatile categories
  • Control: Always overlay with business logic & management assumptions
KPI Trend Analysis
  • What AI does: Identifies KPI trends, seasonal patterns, inflection points
  • Detects: Outliers, accelerations, decelerations vs. historical norms
  • Drives: Follow-up questions for business owners
Narrative Summaries
  • What AI does: Drafts executive summary of monthly/quarterly results
  • Format: Highlights key drivers, misses, opportunities
  • Caution: Requires 30-50% human editing; verify data & tone
Board-Ready Reporting
  • What AI does: Formats data, creates charts, drafts story/narrative flow
  • Control: CFO must review & own all narratives for board
  • Red flag: If you can't explain a chart, remove it or add human context
Waterfall & Bridge Analysis
  • What AI does: Auto-generates variance bridge from prior period to current
  • Identifies: Largest drivers of change (organic, pricing, FX, M&A)
  • Validates: Always reconcile bridge to GL totals manually

FP&A Data Validation Checklist

Controls
0 of 10 completed

Before AI Analysis

After AI Analysis

Data quality: AI forecast accuracy = input data quality. Garbage in = garbage out. Audit data before forecasting.

Model refresh: Retrain models quarterly; if accuracy drops >10%, investigate & adjust assumptions.

Seasonality: AI needs 24+ months of data to capture seasonality accurately. Use caution with shorter histories.

Business context: AI cannot know about planned initiatives, M&A, or market disruptions. Always layer management assumptions.

Board narratives: AI-generated text should be edited by finance leader; remove jargon & ensure alignment with strategy.

Scenario sensitivity: Document all assumptions (revenue growth %, cost inflation, headcount). AI can vary them; humans validate logic.

Trailing edge: Use forecasts for planning, not as "ground truth". History informs; assumptions drive results.

Top FP&A vendors
AnaplanPlanfulPigmentMosaicAdaptive PlanningCubeDatarailsJirav

AI Capabilities Explained

No jargon. Simple explanations of what makes AI tick in accounting.

OCR (Optical Character Recognition)

Reads printed or handwritten text from images & PDFs. Extracts fields (invoice number, amount, date) from unstructured documents.

In AP: Invoice → text extraction → structured data (vendor, line items, amount)

Machine Learning Models

Systems trained on historical data to recognize patterns & make predictions. Improve automatically as they see more examples.

In AR: Customer payment history → model learns to predict payment likelihood

Predictive Scoring

Assigns a numeric score (0-100) to indicate likelihood of an outcome. Higher score = higher risk/opportunity.

Examples: Customer payment risk (0-100), fraud likelihood, invoice error probability

Anomaly Detection

Identifies data points that deviate significantly from normal patterns. Flags unusual GL balances, transactions, amounts.

In Close: Detects journal entries outside historical range, duplicate entries, unusual intercompany balances

Pattern Recognition

Learns recurring sequences in data. Identifies similar items for matching (invoice-to-PO), clustering (vendor categorization), classification (expense category).

In AP: Matches invoice line item → PO line → receipt. Detects duplicate vendor entries.

Generative AI (LLMs)

Large Language Models trained on massive text datasets. Understand & generate human language. Can follow instructions, summarize, draft, explain.

In Accounting: Variance explanations, email drafting, policy writing, variance explanation narratives, code documentation

Workflow Automation

Rules engines that execute if-then logic. Route, approve, escalate, or trigger actions based on conditions.

In AP/AR: Route invoice to approver based on amount/vendor, escalate overdue payments, send reminders

Time Series Analysis

Models data over time to identify trends, seasonality, growth rates, inflection points.

In FP&A: Revenue forecast, expense trends, cash flow prediction, KPI trajectory analysis

🧠The common thread
All AI works by: learn from past data → apply learned patterns → predict/suggest future actions. Always verify outputs.

Governance, Controls & Risk Management

How to deploy AI responsibly. Controls framework, policies, red flags, audit trails.

Human-in-the-Loop Design
  • AI suggests; humans decide on material transactions
  • Define $ thresholds (e.g., invoices >$10K require manual approval)
  • Override capability mandatory for all AI recommendations
  • Log all overrides for trend analysis
Segregation of Duties
  • AI requester ≠ AI approver ≠ payment authorizer
  • Close owner ≠ variance explainer reviewer
  • Map roles to workflows & validate in system controls
  • SAOx / audit testing must include AI-assisted processes
Audit Trails & Documentation
  • Log all AI outputs: prompt, timestamp, user, decision, override
  • Variance explanation drafts must show AI version + human edits
  • Reconciliation exception investigations: log decision & evidence
  • Retain logs for 7-10 years (per statute)
Prompt Documentation
  • Document all system prompts used for AI analysis
  • Version control prompts; track changes (what changed, when, why)
  • Publish approved prompts to team; prevent ad-hoc workarounds
  • Archive old prompts; audit trail if disputes arise
AI Usage Policy Guidelines
  • Approved tools & approved use cases only
  • No PII, confidential data, or bank account details in prompts
  • Data residency compliance (where data stored, who can access)
  • Consequence for unapproved AI use (retraining, escalation)
Data Privacy Boundaries
  • Mask SSNs, bank accounts, customer names in AI inputs
  • Use entity IDs or reference numbers instead of PII
  • Restrict AI access to only needed GL accounts/cost centers
  • Never store confidential data in AI vendor systems without legal review
Red Flag Scenarios
  • AI results contradict known business facts → investigate immediately
  • Consistency drop in prediction accuracy → retrain or pause model
  • Unexplained variance explanations → remove from draft until fixed
  • Repeated same override on same rule → rules need adjustment
What NOT to Automate
  • Journal posting (AI suggests; humans approve)
  • Fraud investigations (AI flags; humans investigate)
  • Accounting judgment (AI informs; humans decide)
  • Communication with auditors (humans own, AI supports)

Governance Self-Assessment Checklist

Controls
0 of 10 completed

Strategy & Oversight

Execution & Monitoring

Purpose: Define responsible use of AI tools in accounting operations. Ensure controls, compliance, & audit readiness.

Approved Tools: [List specific tools by category, e.g., Bill for AP, FloQast for close, ChatGPT for drafting]

Approved Use Cases:

  • AP: Invoice extraction, 3-way match, approval routing, fraud flagging
  • AR: Payment prediction, collection drafts, dispute categorization, risk scoring
  • Close: Variance explanation drafting, anomaly detection, reconciliation, exception triage
  • FP&A: Forecasting, scenario modeling, narrative summaries
  • General: Email drafting, policy writing, data analysis, CSV review

Prohibited Use Cases:

  • Posting journal entries without human approval
  • Making accounting judgment calls (revenue recognition, fair value, capitalization)
  • Sending communications to external parties (customers, vendors, auditors) without review
  • Using confidential data (PII, bank accounts, contracts) without legal approval
  • Unapproved tools or modifications to approved tools

Data Security:

  • No PII (SSN, bank accounts, passport numbers) in AI prompts
  • Use entity IDs, reference numbers, or masked values instead
  • Sensitive data (customer list, contract terms) requires legal review before AI use
  • No storage of data in AI vendor systems without data processing agreement

Audit & Documentation:

  • All AI prompts versioned & stored in repository (Confluence, SharePoint, etc.)
  • Audit trail required: user, timestamp, prompt, output, decision, override (if any)
  • Monthly review of AI-assisted transactions (5-10% sample) by finance manager
  • Quarterly metrics: accuracy, override rate, cycle time improvement

Training & Compliance:

  • Annual training for all users on approved tools & policy
  • Role-based training (AP users vs. close accountants vs. FP&A analysts)
  • Non-compliance consequences: retraining (first offense), escalation (repeat)

Review & Approval: CFO (sponsor), Finance AI committee, Legal, IT Security

🛡Golden rule
If you can't explain it, don't use it. AI enables efficiency; controls enable trust.

AI Prompt Library for Accounting

100 ready-to-use prompts across 10 categories. Copy, paste, adapt to your data. Always review outputs before using.

Prompts for every stage of the close — from journal entries to flux analysis to final review. Copy, paste your data, and get a working first draft.

Variance Explanation Draft
You are an accounting analyst. Draft a 6-8 sentence variance explanation for the month. Format as narrative paragraph.

Input data:
[Paste your KPI table or variance summary here]

Instructions:
- Summarize the largest variances (vs. budget or prior year)
- Identify drivers (volume, pricing, cost, mix)
- Note any one-time items that distort comparison
- Use cautious language ("may indicate", "appears driven by")
- Do NOT invent data or numbers
- Conclude with 3 follow-up questions for business owner

Format: One paragraph, clear, CFO-ready.
Reconciliation Exception Triage
You are an accounting controller. Triage these reconciliation exceptions into categories.

Exception list:
[Paste reconciliation exceptions here - use IDs, not customer names]

Categorize each as:
1) Timing difference (expected to clear next period)
2) Missing documentation (need to retrieve supporting doc)
3) Transposition error (data entry mistake)
4) Potential fraud flag (investigate immediately)

For EACH exception, recommend:
- Next action (follow-up with whom)
- Owner / investigator
- Evidence needed

Format clearly. Prioritize fraud flags at top.
Close Process Gap Analysis
You are a finance process analyst. Analyze our close process for automation opportunities.

Current process:
[Describe current close steps: who does what, timeline, pain points]

Example format:
- Step 1: GL cutoff (2 days, manual, error-prone)
- Step 2: Reconcile balance sheet accounts (5 days, spreadsheet-based)
- Step 3: Variance explanations (3 days, drafted in Word)
- Step 4: Review & approval (1 day, email handoff)

For EACH step, recommend:
- Automation opportunity (AI, RPA, tool)
- Estimated time saved
- Implementation complexity (low/med/high)
- Required controls

Prioritize highest impact + lowest complexity.
Accrual Completeness Check
You are a senior accountant. Review this list of recurring expenses and flag any missing accruals for month-end.

Expense list:
[Paste your recurring vendor list, contracts, or prior month accrual schedule]

For each item, check:
- Was an invoice received this period? (yes/no/unknown)
- If no invoice: estimate accrual amount based on prior months or contract terms
- Flag any amounts that changed >15% from prior month
- Note any new vendors or contracts not yet in the accrual schedule

Output: Table with columns: Vendor | Category | Prior Month | This Month Estimate | Variance | Action Needed

Tone: Precise. Flag uncertainties clearly.
Journal Entry Reviewer
You are an internal auditor. Review these manual journal entries for potential issues.

Journal entries:
[Paste JE data: date, preparer, account, debit, credit, description]

Check for:
1) Round-number entries (e.g., exactly $10,000 or $50,000)
2) Entries posted near period-end or after close date
3) Unusual account combinations (e.g., revenue + expense in same entry)
4) Entries with vague descriptions ("adjustment", "correction", "misc")
5) Entries by users who don't normally post to these accounts
6) Debit/credit imbalances

For each flagged entry, explain WHY it's flagged and suggest follow-up.

Format: Risk-ranked list. Highest risk first.
Balance Sheet Flux Analysis
You are a financial analyst. Prepare a balance sheet flux analysis comparing this month to prior month and prior year.

Balance sheet data:
[Paste: account name, current month balance, prior month balance, prior year balance]

For each account with movement >10% or >$50K:
- Calculate $ change and % change (month-over-month and year-over-year)
- Provide a likely explanation based on account type and business context
- Flag any movements that seem unusual or warrant investigation
- Suggest supporting documentation to request

Format: Table with narrative explanations. Group by: Current Assets, Non-Current Assets, Liabilities, Equity.
Intercompany Reconciliation
You are an intercompany accountant. Reconcile these intercompany balances and identify discrepancies.

Intercompany data:
[Paste: Entity A balance, Entity B balance, transaction type, period]

For each intercompany pair:
- Compare reciprocal balances (Entity A's receivable from B vs. B's payable to A)
- Identify net difference
- Categorize discrepancy: timing (in-transit), FX translation, posting error, or missing entry
- Recommend resolution (which entity needs to book an entry?)

Also flag:
- Any pair with difference >$5K or >5%
- Balances that have been unresolved for 2+ months
- One-sided entries (exists in one entity but not the other)

Format: Reconciliation table with action items column.
Prepaid & Deferred Schedule Review
You are a staff accountant. Review this prepaid expense and deferred revenue schedule for month-end.

Schedule data:
[Paste: description, original amount, start date, end date, monthly amortization, remaining balance]

Check for:
1) Amortization accuracy (does monthly amount x remaining months = remaining balance?)
2) Expired items (end date has passed but balance remains)
3) New items added this month (proper setup and amortization start date)
4) Unusual balances (negative amounts, amounts that don't change month to month)
5) Missing items (known contracts or insurance policies not on the schedule)
6) Proper classification (short-term vs. long-term split for balance sheet)

Produce:
- Summary of total prepaid/deferred balances by category
- List of exceptions found
- Recommended journal entries to correct any issues

Format: Exception report with recommended actions.
Fixed Asset Roll-Forward
You are a fixed asset accountant. Prepare a fixed asset roll-forward for the period.

Asset data:
[Paste: asset description, category, acquisition date, cost, accumulated depreciation, net book value, useful life, method]

Produce:
1) Opening balance reconciliation (beginning NBV by category)
2) Additions this period (new assets placed in service)
3) Disposals / retirements (assets removed, gain/loss calculation)
4) Depreciation expense (current period, by category)
5) Ending balance (closing NBV by category)
6) Reconciliation check (opening + additions - disposals - depreciation = ending)

Also flag:
- Fully depreciated assets still in service (confirm they still exist)
- Assets with unusual useful life assumptions
- Any assets with $0 salvage value and high original cost
- Impairment indicators (significant decline in use or market value)

Format: Roll-forward table with narrative notes.
Revenue Recognition Checklist
You are a revenue accountant. Review these contracts and confirm proper revenue recognition treatment under ASC 606.

Contract data:
[Paste: customer, contract value, deliverables, payment terms, start/end dates]

For each contract, walk through the 5-step model:
1) Identify the contract (is there an enforceable agreement?)
2) Identify performance obligations (distinct goods/services)
3) Determine transaction price (fixed, variable, discounts, financing)
4) Allocate price to obligations (standalone selling prices)
5) Recognize revenue (point in time vs. over time, and why)

Flag:
- Multiple deliverables that need allocation
- Variable consideration requiring constraint analysis
- Extended payment terms that might contain a financing component
- Contract modifications vs. new contracts
- Bill-and-hold arrangements

Format: Contract-by-contract analysis. Include recommended journal entries.
Bank Reconciliation Prep
You are a cash accountant. Prepare a bank reconciliation for this account.

Bank statement data:
[Paste: date, description, amount, running balance — from bank statement]

GL cash account data:
[Paste: date, description, amount, running balance — from general ledger]

Reconcile:
1) Match transactions between bank and GL (by amount and approximate date)
2) Identify outstanding checks (in GL but not yet cleared at bank)
3) Identify deposits in transit (in GL but not yet on bank statement)
4) Flag bank charges/fees not yet recorded in GL
5) Flag interest income not yet recorded in GL
6) Identify unmatched items on both sides

Produce:
- Bank reconciliation (bank balance + deposits in transit - outstanding checks = GL balance)
- List of reconciling items with recommended journal entries
- Aged outstanding items (anything >30 days = investigate)

Format: Standard bank reconciliation format.
Close Status Email Draft
You are a Controller. Draft a close status update email for the finance team and leadership.

Close data:
[Paste: close checklist with completion status, key metrics, open items]

Current close day: [Day X of Y]

Include:
1) Overall status (on track / at risk / behind — and why)
2) Completed milestones (what's done)
3) In-progress items (who owns them, expected completion)
4) Blockers (what's holding things up, who needs to act)
5) Key financial highlights (preliminary revenue, expenses, net income vs. prior month)
6) Open action items with owners and deadlines
7) Next 24-hour priorities

Tone: Concise, factual, no fluff. CFO should be able to scan this in 60 seconds.
Format: Short email — max 15 lines. Use bold for key numbers and action items.
Lease Accounting Entry Prep
You are a senior accountant. Prepare the monthly lease accounting entries under ASC 842.

Lease data:
[Paste: lease description, commencement date, term, monthly payment, discount rate, classification (operating/finance)]

For each lease, calculate and prepare:
1) Monthly amortization of right-of-use (ROU) asset
2) Monthly interest on lease liability (finance leases)
3) Monthly straight-line lease expense (operating leases)
4) Lease liability balance roll-forward (beginning + interest - payment = ending)
5) ROU asset balance roll-forward (beginning - amortization = ending)
6) Short-term lease expense (leases under 12 months, if elected)

Also check:
- Any leases with modifications this month (remeasurement needed?)
- Any leases expiring within 90 days (renewal decision needed?)
- Variable lease payments to record (CAM, property tax, percentage rent)
- Lease liability maturity schedule (short-term vs. long-term split for balance sheet)

Format: Journal entry detail + roll-forward schedules + balance sheet classification.

What prompt is working for your team?

Share a prompt that has saved you time or improved your output. We review submissions and add the best ones to this library.

💬Prompt hygiene
Always mask PII. Review AI output before using. Document prompts in repository. Retrain team on updates.

30-60-90 Day AI Implementation Plan

Phased rollout. Build foundation, expand scope, scale governance. Realistic timeline, measurable outcomes.

Implementation Timeline

1Days 1-30: Foundation
  • Assign AI champion (finance manager with tech interest)
  • Pick 1 pilot workflow (e.g., AP invoice processing)
  • Establish baseline KPIs (cycle time, error rate, cost/transaction)
  • Evaluate 2-3 tools; run proof-of-concept
  • Document current process & define review controls
  • Train 5-10 power users on approved tool
  • Deploy to pilot; monitor daily for first 2 weeks
2Days 31-60: Expand
  • Roll out pilot to full AP team (or first workflow team)
  • Launch 2nd workflow (e.g., close variance explanations)
  • Integrate tool with ERP / close system (if possible)
  • Measure KPI progress vs. baseline; adjust if needed
  • Document lessons learned; refine controls
  • Create prompt library; publish to team
  • Brief audit on AI controls & audit trail setup
3Days 61-90: Standardize
  • Expand to 3rd workflow (e.g., AR collections or FP&A forecast)
  • Finalize & publish AI usage policy
  • Establish governance framework (finance AI committee, roles)
  • Create SOP docs; train full team on approved workflows
  • Measure total time/cost savings; document ROI
  • Present results to leadership; plan next wave (new tools, workflows)
  • Audit sign-off on controls & 1st cycle proof

Implementation Success Metrics

Measurement
0 of 13 completed

30-Day Targets

60-Day Targets

90-Day Targets

Week 1: Kick-off email from CFO. Announce AI initiative, pilot workflow, champion name. Explain benefits & address concerns.

Week 2-4: Weekly 30-min team sync. Demo tool, answer Q&A, celebrate early wins. Publish tips & tricks.

Day 30: 30-day review presentation. Show metrics (time saved, accuracy, team feedback). Answer what went wrong & plan fixes.

Days 31-60: Bi-weekly syncs. Launch 2nd workflow. Publish SOP docs & prompt library. Normalize AI in daily work.

Day 60: 60-day business review. Present to leadership (CFO, CEO, audit committee). ROI, next steps, budget for wave 2.

Days 61-90: Monthly syncs. 3rd workflow launch. Policy finalization. Team training on governance.

Day 90: 90-day celebration & planning. Announce results, recognize team, unveil year 2 roadmap.

📅Realistic pace
90 days for 3 workflows + governance foundation. Don't boil the ocean. Prove value, scale what works.

AI Maturity Model for Finance

Assess your readiness. Define your target state. Plan progression.

Maturity Self-Assessment

Assessment
0 of 16 completed

Organization

Technology & Process

Controls & Compliance

Measurement

🎯Your target state
Most organizations: 18-24 months from Level 1 → Level 3. Maturity is journey, not event.