Customers
The base record for every investor in the book, refreshed alongside everything else.
Customer Segmentation
Your whole book, scored and current — not one client file at a time.
Every investor in the book is re-profiled automatically, every week. Filter the whole roster by behavior, segment, bias, or holdings, and see portfolio health benchmarked against the actual market — without opening a single file to find out who needs attention.
Book snapshotthis Saturday's run
Disposition effect19 clients
Concentration14 clients
Herding9 clients
Overtrading11 clients
Below TLREF8 clients
Data-sparse5 clients
01 / Refresh
A batch that runs itself, every Saturday morning.
No one has to remember to run this. A scheduled job walks the entire brand book on a fixed weekly cadence and writes a fresh snapshot for every investor into a dedicated data layer.
Biases
The behavioral engine's bias read for each client, recomputed from that week's trading and holdings activity.
Bias metrics
The underlying evidence behind each bias call — not just a label, the numbers that produced it.
Holdings
Current portfolio composition, captured as of the run.
Insights
The narrative read-outs generated from that week's profile.
Segments & factor scores
Cohort assignment and factor scoring, so a client's segment reflects this week's behavior, not a stale one.
Always current
Every profile in the listing is at most a week old, refreshed on a fixed schedule rather than whenever someone happens to trigger it.
Book-wide, not per-client
One run touches the entire brand's roster — this is the layer built to look across everyone at once, not deep on one account.
Built to run at scale
Real engineering work went into keeping the batch fast as the book grows — caching, deduplication, and a data-access pattern designed for a full-book pass, not a single client.
A report someone has to remember to run gets stale. A batch that runs itself doesn't.
02 / Filter
Slice the whole book, instead of opening files one at a time.
A filterable listing sits on top of the weekly snapshot — advisors, team leads, and compliance reviewers can narrow the entire roster by behavior pattern, segment, specific bias, or holdings composition in one view.
Segmentation listing
Find who needs attention this week, not next quarter.
Instead of opening client files one by one, the listing lets you start from a question — who's showing disposition effect, who's concentrated in a single name, who's below TLREF this month — and narrows the book to exactly those clients. Built for the people who own dozens or hundreds of accounts, not one.
Filter by behavior pattern, segment, bias, or holdings
Built for advisors, team leads, and compliance
Jump straight into any client's snapshot from the list
Listing filterswhole book · live
1 Pick a dimension segment, bias, holdings, behavior
2 Book narrows to matching clients no manual lookup
3 Open any client's snapshot directly one click in
4 Combine filters for a compliance sweep whole-book review
The question isn't "how is this one client doing." It's "who, across everyone, needs me this week."
03 / Health
Health scored against the market, not a vibe.
Every snapshot benchmarks portfolio health against real market context — a Sharpe ratio, a comparison against TLREF, and a return adjusted for actual published inflation, not a rough estimate.
Portfolio Health
Three numbers, each benchmarked against something real.
A Sharpe ratio computed from the client's own return series. A comparison against TLREF, the reference funding rate, so "beating the market" has an actual funding-cost baseline. And a real inflation-adjusted return, anchored to the latest officially published CPI print — because CPI data itself lags roughly five months before publication, and this benchmark is honest about that lag instead of faking a same-day inflation figure.
Sharpe ratio, computed per client
Benchmarked against TLREF, the reference funding rate
Inflation-adjusted return, anchored to the latest published CPI
Health snapshotper client · weekly
1 Sharpe ratio from the real return series risk-adjusted
2 Compared against TLREF funding-rate baseline
3 Adjusted for real, published CPI latest print, not same-day
4 Too little history to score? flagged, not faked
A client with too little trading history simply shows as data-sparse rather than getting a fabricated score — the system doesn't force a rating onto an account that doesn't have enough activity to support one.
A client with two trades this month isn't rated. They're flagged as too early to say — which is the honest answer.
Depth
Breadth here. Depth lives one layer down.
Customer Segmentation answers "across my entire book, who needs attention, and how are they doing." A separate app, Customer Profiling, goes deep on a single client the same behavioral engine powers both — this is the book-wide, always-current view; that one is the one-client, on-demand deep dive.
Start here to find who needs a look. Open a single client's full profile when you're ready to go deep on that one account.
Breadth tells you where to look. Depth tells you why.
Whole book, refreshed weekly
Customer Segmentation re-profiles every investor in the brand automatically, on a fixed schedule.
One client, on demand
Customer Profiling goes deep on a single account when an advisor is working that specific relationship.
Same engine underneath
Both apps read from the same behavioral engine — the difference is scope and cadence, not the underlying analysis.
Filter first, then dive
The natural path is to filter the book here, then open the specific client in Profiling for the full picture.
04 / FAQ
The questions a team lead asks before trusting a batch job.
Short, specific answers — the same understated register as the rest of this platform.
How current is "weekly," really?
The batch runs Saturday mornings, automatically, across the entire brand book. Every snapshot in the listing is at most a week old — there's no manual step for anyone to skip or forget, so "current" doesn't depend on someone remembering to click run.
What happens to a brand-new client with no trading history yet?
They show up as data-sparse rather than getting a fabricated score. The system doesn't force a Sharpe ratio or a bias read onto an account that doesn't have enough activity to support one — it says so instead.
Can compliance use this for a book-wide review?
Yes — the same filterable listing that an advisor uses to find who needs attention lets a compliance reviewer slice the whole book by segment, bias, or holdings composition in one pass, instead of opening client files individually.
Does this replace Customer Profiling?
No — they're complementary. Segmentation is the book-wide, always-current layer for deciding where to look first; Profiling is the single-client, on-demand deep dive for once you've decided which account to work.
Why TLREF and CPI specifically for the health benchmark?
TLREF is the reference funding rate, so a client's return has an actual cost-of-capital baseline to clear, not an arbitrary hurdle. The inflation-adjusted figure uses real published CPI, anchored to the latest officially released print, because CPI itself lags roughly five months — the benchmark is upfront about that lag rather than estimating around it.
See your own book, filtered and benchmarked, in the first walkthrough.
Request a walkthrough focused on your brand's book — the segments, biases, and health benchmarks it would surface in week one.