EBITDA Normalization & Add-Back Automation
Upload your GL and let Shepi surface add-back candidates automatically. Every transaction analyzed, every adjustment categorized, every finding documented.
100%
GL transactions scanned
5 categories
Adjustment taxonomy
Minutes
To first adjustment candidates
What Is EBITDA Automation?
EBITDA normalization is the process of adjusting reported earnings to reflect the true, recurring earning power of a business. Traditionally, this means an analyst manually reviewing every account, every vendor, and every transaction to identify items that don't represent ongoing operations.
Shepi automates this by scanning 100% of GL transactions using AI pattern recognition, categorizing potential adjustments using a standardized taxonomy, and presenting findings for your review — with supporting evidence attached.
What Gets Automated
GL scanning
Every transaction reviewed for adjustment candidates — not a sample, the entire ledger
Pattern recognition
AI identifies recurring vs non-recurring items, seasonal patterns, and anomalies
Categorization
Each candidate automatically classified: owner/seller, non-recurring, pro forma, related party, or accounting policy
Quantification
Dollar impact calculated across all analysis periods with period-over-period trending
Documentation scaffolding
Each adjustment linked to source transactions with suggested rationale
Bridge generation
Net income to adjusted EBITDA bridge built automatically as adjustments are confirmed
Adjustment Types Detected
Owner compensation
Above-market salary, bonuses, benefits, personal expenses — normalized to market-rate replacement
Non-recurring expenses
Litigation, one-time consulting, disaster costs, COVID impacts, restructuring charges
Non-recurring revenue
PPP forgiveness, insurance proceeds, asset sales, one-time contract revenue
Pro forma adjustments
Annualized contracts, mid-period price changes, new hires/terminations
Related party transactions
Below/above-market rent, management fees, non-arm's-length vendor arrangements
Accounting policy
Revenue recognition timing, capitalization vs expensing, accrual differences
How It Works
Ingest financial data
Connect QuickBooks, upload trial balance, or import GL export
AI scans the ledger
Pattern recognition identifies adjustment candidates across all accounts and periods
Review candidates
Each finding presented with category, amount, confidence score, and source transactions
Apply judgment
Accept, modify, or reject — you control the final adjusted EBITDA
Export the bridge
Generate a formatted EBITDA bridge with supporting detail for deal parties
Automated vs Manual EBITDA Normalization
| Dimension | Manual Process | Shepi Automation |
|---|---|---|
| Coverage | Sample-based review | 100% of transactions |
| Time to first findings | 1–2 weeks | Minutes |
| Categorization | Analyst judgment only | AI + analyst judgment |
| Consistency | Varies by engagement | Standardized taxonomy |
| Documentation | Built manually in Excel | Auto-generated with source links |
| Period comparison | Separate analysis per period | Multi-period analysis built-in |
Use Cases
Pre-LOI screening
Run a quick EBITDA normalization before committing to a deal — know what you're buying
Sell-side preparation
Identify and document add-backs before buyers do — control the narrative
CPA engagement acceleration
Start with AI-identified candidates, let the CPA focus on judgment calls
Portfolio monitoring
Track normalized EBITDA across portfolio companies with consistent methodology