Like having a senior acquisitions partner reviewing every deal.
Institutional-grade AI analysis on every commercial real estate deal you underwrite. Tampa multifamily. Lexington industrial NNN. Whatever the asset — get the underwriter perspective that finds what spreadsheets miss.
Sample shown below is real AI output from a sample deal in UpsideIQ — not marketing copy.
Live AI analysis on a sample deal.
Example output on a 50-unit Class B multifamily in Tampa, FL ($5.0M, $100k/unit). Your analysis will cite specific numbers from your own deal data.
Below 1.0× the property can't cover its debt; lenders typically want 1.25×+.
Equity out in Y0, annual distributions, sale-laden exit.
AI Deal Analysis
- • 50-unit multifamily in Tampa, FL at $5.0M ($100k/unit)
- • 70% LTV financing at 6.50% with 3-year I/O, 30-year amortization
- • Year 1 NOI $251,305 (5.08% actual cap rate on purchase price)
- • 5-year hold with planned refi at 5.50% cap, exit at 7.00% cap
- • LP equity required $1.4M; projected exit value $4.29M (14% below basis)
- • DSCR 0.95x on P&I (below lender minimums); LP returns deeply negative (-21% IRR, 0.31x EM)
- • Tampa multifamily market has strong fundamentals with population and job growth tailwinds
- • NOI grows from $251k Y1 to $300k Y5 (19% cumulative growth) suggesting some organic rent growth potential
- • Basis at $100k/unit is reasonable for Tampa depending on submarket and vintage
- • 3-year I/O period provides initial cash flow cushion during lease-up or repositioning phase
- • DSCR at 0.95x is materially below lender minimums (1.20-1.30x); this loan likely won't fund without significant equity cure or rate buydown
- • Exit valuation at $4.29M is 14% below $5.0M purchase price after 5 years, indicating negative value creation
- • LP equity multiple of 0.31x means investors lose 69% of capital; IRR of -21% is catastrophic
- • Exit cap rate at 7.00% vs entry at 5.03% implies 197bps of cap rate expansion—aggressive assumption that market deteriorates significantly
- • Refi proceeds are negative $440k, meaning additional equity injection required at Year 5 to execute refinance
- • Operating expenses at $327k on $251k NOI suggest 56.5% expense ratio (NOI margin only 43.5%), which is high for multifamily
The engine's base case already shows structural failure: 0.95x DSCR won't clear lender underwriting, -21% IRR and 0.31x EM are non-starters for any LP. The exit cap rate assumption of 7.00% vs 5.03% entry implies severe market dislocation. Even with NOI growth to $300k by Year 5, the exit value of $4.29M destroys $710k in equity value. The refi scenario at 5.50% cap produces negative proceeds of $440k, requiring additional capital injection. This deal is fundamentally mispriced at $5.0M purchase.
In stress, reduce Y1 NOI by 10% to $226k (DSCR falls to 0.86x, further below lender minimums). Push exit cap to 7.50% (+50bps from base 7.00%), driving exit value to ~$4.0M and LP EM to ~0.25x. Extend lease-up by 6 months, compressing stabilization NOI. The deal fails in base case, so stress case is academic—this is a structural pass. Even if we assumed a more realistic 6.00% exit cap (still 97bps above entry), exit value would be ~$5.0M, producing flat equity returns before fees and carry.
- What is the actual in-place occupancy and current effective rent per unit? Y1 NOI of $251k seems low for 50 units.
- Why is the exit cap rate modeled 197bps higher than entry? What market deterioration scenario justifies this assumption?
- What is the business plan to bridge the 0.95x DSCR to lender-acceptable levels? Is there a rate buydown or mezzanine equity contemplated?
- Can you provide T12 operating statements? The 56.5% expense ratio suggests either deferred maintenance, high property taxes, or operational inefficiency.
- What comparable sales support a $100k/unit basis in this Tampa submarket? Recent comps and submarket specifics needed.
This deal is a structural pass. The purchase price of $5.0M produces a sub-1.0x DSCR that won't clear any lender, and the exit assumptions destroy 69% of LP equity over five years. The 197bps cap rate expansion from entry to exit is either a modeling error or reflects an expectation of severe market dislocation that would make the acquisition timing catastrophically wrong. Even in a best-case scenario with more realistic exit cap assumptions, the deal produces flat-to-negative returns. This needs to be repriced at $4.0-4.2M to generate institutional returns, or we walk.
This is actual output from UpsideIQ's AI analysis on a 50-unit Class B multifamily deal in Tampa, FL ($5.0M, $100k/unit). Your analysis will cite specific numbers from YOUR deal data — not generic templates.
Not a number-cruncher. A second opinion that listens.
The OM tells the AI what the broker wants you to know. The numbers tell it what the engine computed. But context — seller motivation, off-market relationship, portfolio fit — only you know. Add it; the AI re-analyzes around your situation.
Concern
Sub-1.0 DSCR (0.95×) on a 50-unit Tampa MF at $5M basis. The deal cannot cover its own debt service from operations at the modeled exit cap.
Overall
2/10 — Structural pass. Walk unless the broker materially reprices.
Updated view
Estate-driven motivation + off-market access materially changes the picture. The 0.95× DSCR at ask is unworkable, but a 10-15% basis reduction is realistic without competing bids. At ~$4.3M the deal pencils.
Recommended input change
Purchase Price → $4,300,000
Rationale: estate motivation + off-market = no competing bids; standard 10-15% off-market discount applies; resulting DSCR clears 1.10×.
Overall
6/10 at $4.3M basis. Submit at $4.2M with a 30-day close.
Same engine. Same numbers. Your context.
Every regeneration runs against your saved deal state plus the engine's output. The AI's view evolves with the situation — and every version is preserved so you can show partners and LPs exactly how the analysis developed. Cached; no per-view cost; counts against your daily AI cap only when you refine.
From deal entry to AI critique in under a minute.
Underwrite your deal in UpsideIQ
Enter deal details or upload an OM. The engine runs a full 10-year DCF, computes IRR / EM / DSCR, and surfaces rules-based health flags.
Click Generate AI Analysis
One click on the preview page sends your deal data + engine outputs to Claude with strict anti-hallucination guardrails.
Get institutional critique in ~30s
Structured output: key facts, what to like, what to worry about, base + stress case, 6-dimension scoring, and broker questions to ask.
Built like an analyst, not a chatbot.
Cites specific numbers from YOUR deal
Every claim references actual values from your inputs and engine outputs. "DSCR at 0.95x" — not "DSCR looks weak."
Identifies structural risks
Negative leverage. DSCR shortfalls. Cap rate expansion that destroys equity. Refi proceeds that go negative. The things that fail deals at IC.
Specific broker questions, not generic checklists
Each analysis ends with 3-5 questions to ask the broker — tied to actual data gaps and assumption stress points on YOUR deal.
Scored across 6 dimensions
Basis, Building, Market, Tenant / Demand, Lease / Rent Structure, Upside — each 0-10. Aggregate overall score 0-10.
Anti-hallucination guardrails
Strict system prompt: only state facts present in deal data. Banned phrases scanned post-generation. Confidence rating self-assessed.
Cached intelligently
Re-views of unchanged deals reuse the cached analysis — no waste on cosmetic edits. Regeneration only on material input change.
Built on a verified underwriting engine.
AI analyzes the engine output. The engine itself is the load-bearing math — verified across 30+ test scenarios with hard invariants and bit-identical canary baselines.
MF + Industrial scenarios run on every engine-touching commit. Hard invariants on valuations, DSCR, IRR finite-ness.
Engine validated line-by-line against canonical Excel underwriting models. Documented divergences in append-only ledger.
Anti-hallucination system prompt + post-generation phrase scanning. AI references only data we provide.
Included on every Pro plan.
No add-ons. No metered API fees. Unlimited deal analyses within your daily generation cap.
AI analysis · Unlimited deals
Save 29% vs monthly
14-day money-back guarantee on annual plans
Frequently asked questions
Is my deal data sent to external AI?
Yes — analysis uses Anthropic's Claude API. Your deal data is processed under Anthropic's enterprise data handling policies. Cached results stay in UpsideIQ; only fresh generations send new data. A one-time privacy disclosure modal appears on your first generation.
Can I disable AI analysis?
Yes — AI generation is opt-in per deal. You manually click "Generate AI Deal Analysis" on the preview page when you want it. Nothing runs in the background. PDF reports include AI analysis only if you toggle it on (default OFF).
How much does it cost?
Included on Pro plans at no additional charge, up to 10 generations per user per day. Typical analyst usage is 1-2 fresh analyses per day; cached re-views of unchanged deals are free.
What if the AI makes a mistake?
AI analysis is supplementary — always verify claims independently. UpsideIQ includes anti-hallucination guardrails (system prompt forbids inventing facts, post-generation phrase scanning, self-assessed confidence rating) but AI output is not investment advice.
How is this different from asking ChatGPT?
AI analysis runs against the engine's structured deal data: NOI, DSCR, IRR, EM, projected exit, refi math, anomaly flags, peer benchmarks. ChatGPT has no context for any of that. Here, the AI cites specific numbers from your deal — and the anti-hallucination rules prevent it from inventing things.
Which Claude model powers this?
Claude Sonnet 4.5 — Anthropic's balanced quality / cost model. The system is designed to swap to a newer model when available without breaking the institutional voice or anti-hallucination guardrails.
Run AI analysis on your next deal.
Free to start. Pro plans include unlimited deal analyses.