Skip to main content
Industries · Banking & Wealth

Autonomous AI discovery built for banking & wealth.

Continuous AI discovery across your deposit, advisory, portfolio, and client-transaction data. Every finding ships with the underlying queries, the cohort definition, and the cost of the run. The "where is the book actually moving" question, answered before your next portfolio review.

Built for Heads of Retail Banking, Heads of Wealth Management, Chief Data Officers, and Risk & Analytics leads at retail banks, private banks, and wealth firms.

CRITICALCash migration
HYSA balances above $250K drained 31% in Q1 while aggregate deposits looked flat.
Agent flagged the shape week 3 of the trend. 1,840…
From a real DecisionBox run
Why DecisionBox

The book-of-business questions nobody had time to query, running overnight against your data.

Your team owns the relationships, the portfolios, the regulatory posture. The agent runs unattended against your deposit, advisory, transaction, and client data — and ships findings into a queue your team triages. Every claim carries the underlying queries and the data-validated number. The agent reads where your data lives; it never writes back, never bypasses RBAC, and logs every step for the audit.

See how the agent reads cross-product behavior
The cost of late signals

The aggregate looks stable. The cohort underneath is where the book actually moved.

Three benchmarks every banking and wealth leader recognizes. Each points to the same gap: your data already holds the signals — pre-attrition behavior, deposit migration, segment-level fee shifts — but they only surface when somebody asks the right question. Most weeks, nobody does.

Advisor-transition attrition
20%
Of wealth clients leave within 12 months of an advisor transition. The aggregate retention number hides the advisor-departure cliff that costs the firm decades of compounded fees.
Cerulli Associates, U.S. Advisor Metrics
Single-product customers
42%
Of bank customers hold only one product with their primary bank. Cross-sell penetration is the slowest-moving lever in retail banking — and the data already has the signals.
J.D. Power, U.S. Retail Banking Satisfaction Study
Revenue concentration
80 / 20
Roughly 80% of wealth firm fees concentrate in the top 20% of clients. When the top segment shifts (advisor change, generational transition, market loss), the impact runs ahead of AUM movement.
McKinsey, North American Wealth Management Survey
From real runs

Three findings the agent surfaced, paired with the recommendation it shipped.

Three shapes the book sees every quarter: a single-product cash migration the deposit aggregate concealed, a cross-product pre-attrition signal that lives across product lines, and a segment-level fee shift the AUM total averaged away. Same auditable trail every time — severity, the underlying queries, the cohort definition, the cost of the run.

CRITICALCash migration
HYSA balances above the FDIC threshold drained 31% in Q1.
period: Q1 (Jan – Mar) · affected clients (HYSA >…
Cross-checked against deposits.hysa_daily…
single-product cash exit
P1Recommendation
Re-price tier + relationship-priced outreach
Re-price HYSA tiers above $250K to close the gap to Treasury;
Expected impact
target: stem high-balance flight; defend net interest…
Target: Re-price tier +…2 concrete actions
CRITICALPre-attrition signal
47 wealth clients drew down checking balances 60+ days before terminating advisory.
pattern: checking drawdown precedes advisory…
Cross-checked across deposits.checking_daily…
cross-system pattern
P1Recommendation
Senior-advisor retention calls + recurring instrumentation
Route the 47 affected clients to senior-advisor retention calls this week;
Expected impact
target: intercept terminations before paperwork is filed…
Target: Senior-advisor…2 concrete actions
HIGHSegment masking
Total AUM held flat WoW.
total AUM (WoW): flat (~$24.8B) · HNW segment…
Cross-checked across portfolio.holdings_monthly…
aggregate hides the segment
P1Recommendation
Orphan re-pairing + board-materials flag
Audit the Q4 advisor transitions;
Expected impact
target: defend HNW fee revenue; prevent further drift…
Target: Orphan re-pairing +…3 concrete actions

Point the agent at your book of business for a week. Triage what it finds Monday morning.

Use cases

Five surfaces the team already owns, one continuous discovery.

Same engine, same data, five categories of work your team is already doing — or wants to do, if the queue ever clears. Pick the surface that matters this quarter.

The client whose checking went quiet eight weeks before the termination paperwork.

The pre-attrition shape that lives across product systems: checking drawdown, advisory engagement decay, portfolio drift away from house funds, support-touch changes in tone. The agent joins the product tables your RMs and advisors don't routinely read together — and surfaces the clients in the early-warning shape weeks before the termination request lands. The deep dive lives in how the agent decides what to investigate.

Sample analysis areas
  • Cross-product pre-attrition pattern detection (checking + advisory + portfolio)
  • Advisor-transition cohort retention monitoring
  • Behavior-change windows that precede formal exit
  • Segment-specific attrition signals (HNW vs. mass-affluent)
  • Recurring weekly scans with alerting for in-tenure clients
Why DecisionBox

Built for banks and wealth firms who want an analyst that runs overnight, and a record the regulator can read.

Three things every finding from DecisionBox carries. None of them changes how your team works in your core banking system, your portfolio accounting platform, or whatever data tools and BI you already standardized on.

Sees what nobody had time to investigate

The agent picks which cohorts, segments, and cross-product behaviors to investigate from your deposit, advisory, portfolio, and transaction data. Findings ship into a queue your team triages in the morning. The pre-attrition signal stops sitting in the data waiting for someone to think to ask.

Every number you'd defend at the portfolio review

The underlying queries the agent ran, the cohort definition, the data window, the cost of the run, the validation checks. Your team re-runs anything, extends anything, overrides anything. No black boxes, no platform-reported numbers the data can't validate, no "trust the AI."

Reads your data where it lives. Audit-ready.

Self-host on Docker, Helm, or Terraform — or managed cloud on your terms. The agent reads your data where it lives, never writes back, never bypasses RBAC, and logs every query and reasoning step for the audit. Snowflake, BigQuery, Databricks, Redshift, PostgreSQL. Open source, AGPL v3. Built for the regulator, the security review, and the data-residency posture you already enforce.

The next portfolio review, with the segment shifts already mapped.