DealDesk AI — M&A Due Diligence in Minutes

Traditional M&A due diligence takes 4–8 weeks, costs $50K–$500K, and produces decisions that are impossible to trace. DealDesk AI compresses that into minutes — without sacrificing rigor or removing the human from the loop. The system deploys 5 specialized AI agents that coordinate exclusively through Band. Each agent has a single job: @Research analyzes the target company's business model, market size, team quality, financials, and competitive positioning. @Risk evaluates every finding and scores risks LOW, MEDIUM, or HIGH — regulatory exposure, key-person dependency, competitive threats. @Valuation builds a financial model: revenue multiples, comparable transactions, and a recommended entry range with deal terms. @Writer synthesizes all three analyses into a structured VC/PE-grade investment memo — executive summary, market analysis, risk section, valuation, and recommendation. @HumanReview presents the memo to a real human who reads it and makes the final call: Approve, Reject, or Request Changes. No memo ships without human approval. Every agent posts its output to a shared Band chat room using its own Agent API key, @mentioning the next agent in the chain. The result is a fully transparent, auditable pipeline — not a black box. Any stakeholder can join the Band chat and watch the analysis unfold in real time. The project includes a real-time Flask dashboard with Server-Sent Events for live pipeline monitoring, and a GitHub Pages landing page. Tech: Band Agent API · Featherless AI · Qwen2.5-72B-Instruct · Python · Flask · SSE · Human-in-the-Loop