DcisionAI
The Intelligent Enterprise Decision Layer
Decision infrastructure for the regulated enterprise — where every high-stakes choice becomes provable, auditable, and permanently traceable.
Executive Summary
DcisionAI Infrastructure for the Regulated Enterprise
DcisionAI turns plain-English problems into certified optimal decisions through a six-agent pipeline with mathematical audit gates. It delivers the best possible answer in minutes, with every binding constraint and tradeoff explained in business language and dollar terms.
Why PE & RIA first: Fiduciary liability, LP reporting mandates, and EU AI Act enforcement make certified decisions non-optional — not a nice-to-have. We enter where the compliance deadline is the sales cycle.
The Problem
Every regulated organization makes high-stakes decisions under dozens of constraints — without proof of optimality, a defensible audit trail, or the ability to re-run when assumptions change.
Wealth Management & RIA — The Reg BI Trap
74% of wealth management firms already use AI — yet most cannot produce a traceable rationale for the decisions those tools influence. FINRA's 2025 Annual Regulatory Oversight Report explicitly links AI use to Reg BI obligations, and the EU AI Act (enforceable August 2026) classifies investment recommendation AI as high-risk, with fines up to €35M or 7% of global turnover.
Unverifiable AI is not automation. It's liability.
Private Equity — The Allocation Black Box
PE funds make $10M–$100M capital allocation decisions under dozens of constraints — ESG mandates, concentration limits, LP preferred returns, covenant restrictions. The math lives in spreadsheets, the rationale in email threads. ILPA 2026 mandates standardized IRR, TVPI, and MOIC reporting with documented allocation rationale — and 'we used a model' doesn't satisfy LP scrutiny.
When assumptions change, you restart the engagement.
Fund Administration — The Waterfall Black Box
Every PE fund runs a bespoke waterfall — preferred return, GP catch-up, carry split — inside a 500-row Excel model that one analyst built years ago. Manual processes carry a 15–20% error rate, and LPs can be misallocated millions. ILPA 2026 requires documented, auditable carry calculation methodology.
ILPA exam question: "Why did LP A get less than LP B?" The answer can't be a spreadsheet.
Our Solution
DcisionAI: Decision Infrastructure
Describe the problem in plain English. DcisionAI routes it through a six-agent pipeline — Discovery, Research, Planning, Model, Solve, Explain — with mathematical audit gates that verify correctness at every stage before any output reaches the user.
Intent Layer
DcisionAI agents interpret the user's intent, extract every constraint — tax lots, risk profile, ESG mandates, concentration limits — and translate them into a precise mathematical formulation. No ambiguity reaches the solver.
Deterministic Layer
DcisionAI solver agent find the mathematically optimal answer across all constraints simultaneously. Every output carries a Certificate of Optimality — proof that no better solution exists given the constraints. Reg BI audit trail built in.
Explanation Layer
Results return in plain English: which constraints are binding, what each one costs in dollars to relax, and what the tradeoffs are. Ready to present to the client or submit to compliance — in minutes.
Not a recommendation. Not a prediction. A certified optimal decision — with every binding constraint named in business language and every shadow price shown in dollar terms.
Each run compounds: the context graph grows richer with rules, precedents, and tradeoffs, making every future decision faster, more accurate, and auditable.
Competitive Moat
The Compounding Advantage
PROVE
Certificate of Optimality
Mathematical proof that no better answer exists. Certificate of Optimality for every decision. Binding constraints and shadow prices fully exposed.
EXPLAIN
Shadow Price Analysis
Shadow prices reveal exactly which constraints are costing money. Plain-English explanations, sensitivity analysis and what-if scenarios on demand.
TRACE
The Context Graph
Every decision stored in a Context Graph — searchable, permanent institutional memory for GPs, auditors, and regulatory compliance.
Every DcisionAI run compounds. The context graph grows richer with every problem — encoding rules, precedents, and tradeoffs in a form that makes every subsequent decision faster, more accurate, and auditable. It becomes a living record of how the organization actually decides.
The enterprises that will define the next decade are the ones that close the gap — turning every decision into a provable, auditable, improvable outcome.
Use Cases
Where DcisionAI Works
The same infrastructure applies wherever high-stakes decisions meet hard constraints. Every problem has a mathematically optimal answer. DcisionAI finds it, in minutes.
PE Portfolio Pricing
Optimize pricing across 10–30 portcos under margin, competitive, and covenant constraints. Board-ready output with full audit trail.
Capital Allocation & LP Waterfall
Allocate LP capital across 40+ assets under ESG mandates and concentration limits. Model catch-up provisions, preferred returns, and fund recycling. Every tradeoff in dollar terms.
RIA Portfolio Construction
Per-client rebalancing and tax-loss harvesting under IPS constraints and Reg BI best-interest requirements. Full audit trail per client.
Operational Decisions
Nurse scheduling, fleet routing, capex allocation, surety bond issuance — any constrained optimization problem with a defensible answer requirement.
Launch Wedge
PE/VC + RIA: Non-Discretionary Demand
Two regulated verticals. Three certified implementations already live. Compliance deadlines make adoption non-optional.
1
PE / VC — Solving the Allocation Black Box
The math that lives in spreadsheets now lives in a certified, replayable model. Capital allocation, portfolio construction, LP waterfall, exit routing, and fund recycling — each decision carries a Certificate of Optimality and a full audit trail. ILPA 2026 LP scrutiny answered with a replayable formulation, not a spreadsheet.
DICLAF IV — $95M certified.
2
Fund Administration — Replacing the Waterfall Spreadsheet
DecisionAI's waterfall skill takes LP commitment schedules, realized proceeds, and bespoke waterfall structures (preferred return, catch-up, carry) and produces a fully auditable distribution calculation. Every LP allocation is traceable to the exact binding constraint. ILPA exam question — "Why did LP A get less than LP B?" — answered with provenance, not a 500-row Excel.
Carry consolidation and gross-to-net reconciliation built in.
3
Wealth Management & RIA — Closing the Reg BI Gap
Per-client portfolio construction, rebalancing, and tax-loss harvesting under IPS constraints — with a documented best-interest rationale for every decision. SEC 2026 exam question — "Show me your AI decision process for this client" — answered with a full audit trail, not a model disclaimer.
Reg BI · EU AI Act · Per-client audit trail.
Market Timing
Why Now
Four forcing functions are converging to create non-discretionary demand for certified AI decisions. The window for first-mover advantage is closing.
EU AI Act · 2026
Credit scoring, capital allocation, risk assessment: audit trail mandatory. Penalty up to €15M or 3% global turnover.
SEC AI Governance · 2026
Model risk management for investment advisors. Registration risk for firms without documented AI decision processes.
$144.6T AUM · Reg BI
AI-assisted allocation requires a demonstrable best-interest process, per client, at SEC examination. Fiduciary duty is enforceable.
Basel IV · Model Risk
Global banks and asset managers face SR 11-7 / Basel IV model validation requirements. Undocumented AI decision models are a material audit finding — and a board-level liability.
Product Architecture
What We've Built
Four composable layers — each independently valuable, compounding together into a defensible infrastructure stack.
Autonomous Agents
Discovery, Research, Planning, Execution, Explanation — each with typed outputs and clear handoffs between stages.
Domain Skills
Domain-specific optimization skills (VC, supply chain, tax, etc.) selected by a taxonomy + guardrail router.
Solvers & Validators
HiGHS + SCIP solvers wired so agents run real optimization — not just advise. Certifiable in under 90 seconds.
Context Graph
Persistent graph of problems, skills, agents, and outcomes — enabling audit, traceability, and cross-decision learning.
Product Architecture
Systems That Compound Over Time
DcisionAI doesn't just solve problems — it gets better at solving them. Four interlocking systems ensure every run makes the next one faster, more accurate, and harder to replicate.
The Pipeline
Every problem runs end-to-end: plain-English input through six specialized agents to a certified optimal decision. Board-ready output in minutes.
The Error Recovery Chain
The platform self-heals most failures automatically. When a problem genuinely needs human judgment, HITL surfaces the exact gap — so the user fills in what only they know, and the job continues without restarting.
The Learning Flywheel
Every successful run contributes proven optimization patterns back to the platform's domain skill library. The more DcisionAI runs, the sharper its models become — permanently.
Active Memory
Past decision graphs don't sit idle. They actively steer current jobs — surfacing relevant precedents, binding constraints from prior runs, and domain methodology accumulated from real problems, not hypothetical ones.
Competitive Landscape
Competitive Defensibility
DcisionAI is the only platform combining mathematical proof, natural language input, full audit trail, instant re-optimization, and persistent decision memory — at a fraction of consulting cost.
vs. LLMs: They guess; we prove. · vs. Gurobi: They're the calculator; we're the accounting system. · vs. McKinsey: They deliver a PDF; we deliver re-optimizable infrastructure.
Market Opportunity
Market Sizing: TAM / SAM / SOM
All figures built bottom-up: institution count × ACV. Analyst sizing used as cross-check only.
TAM — $16.8B
Math optimization software ($4.2B) + Decision intelligence ($7.1B) + AI compliance infra ($5.5B) · 15–18% CAGR
SAM — $1.39B
PE/VC private credit & funds: $613M · Enterprise operations: $516M · AI compliance (cross-cutting): $263M · Growing to $1.6B by Year 3 as skills expand.
SOM — $3M ARR
Warm channels only · 0.36% of SAM · Conservative: $3M (5 PE funds) · Base: $4.95M (5 PE ecosystems × $650K portco multiplier) · Within Keybanc Y3 benchmark.
Go-to-Market
Go-to-Market Strategy
Sell-To — Direct Buyers
PE Operating Partners
Budget Holder. Mandated to drive EBITDA across 10–30 portcos. Currently spending $500K+ per engagement on consulting. DcisionAI replaces the static deliverable with re-optimizable infrastructure. Land one fund, expand across the portfolio.
PE Fund GPs & CFOs
Budget Holder. Responsible for LP reporting, capital allocation, and waterfall compliance. Fiduciary liability and LP mandates make certified decisions non-optional. DcisionAI is the audit trail for every allocation decision.
RIA Firm Owners & Managing Partners
Budget Holder. Accountable for Reg BI compliance across all client accounts. Per-client audit trail and certified portfolio construction decisions reduce regulatory exposure and differentiate the firm.
Independent RIAs
Direct Buyer. Solo and small-firm advisors who are both practitioner and owner. Reg BI compliance burden falls entirely on them. DcisionAI gives them institutional-grade audit infrastructure at a fraction of enterprise cost — and differentiates their practice.
Sell-With — Channel Partners
PE Fund Administrators
Co-sell Partner. Apex, Citco, and SS&C already sit inside LP reporting workflows. DcisionAI becomes the certified optimization layer embedded in their reporting stack, with a warm referral path to GPs and LPs.
Wealth Tech & RIA Platforms
Integration Partner. Orion, Riskalyze, Tamarac, and similar platforms serve thousands of RIAs. Whitelabel or API integration embeds DcisionAI's portfolio optimization and audit trail directly into existing advisor workflows.
Big 4 & Boutique Advisory Firms
Delivery Partner. Deloitte, EY, and PE-focused boutiques like Accordion and Alvarez & Marsal deliver operational improvement mandates across portcos. DcisionAI replaces the static PDF deliverable with re-optimizable infrastructure, making them stickier and us the engine behind their engagements.
Advisors at Larger RIA Firms
Internal Champion. Senior advisors who feel the compliance and documentation burden daily. They surface the need, pressure firm leadership to act, and become power users once deployed. Bottom-up adoption path into enterprise RIA accounts.
Distribution motions: Partner-Led (PE operating partners + RIA compliance) · Product-Led (Salesforce AppExchange) · Direct Enterprise (pricing pilot wedge → portfolio expansion)
Pricing Model
Decision-Based Pricing Tiers
Pilot — $25K
3 months · 500 decisions
Starter — $100K
Annual · 2,500 decisions
Growth — $250K
Annual ·7,500 decisions
The Future of Decisions is Deterministic.
Wedge
Regulated Decision Infrastructure — entering where compliance deadlines make certified, auditable AI decisions non-optional.
Moat
Mathematical proof. Full audit trail. Instant re-optimization. Built into the decision — not bolted on after.
Scale
Horizontal Decision Infrastructure — from PE portfolios to corporate strategy, government, and healthcare.
DcisionAI turns high-stakes decisions into provable, auditable institutional assets.