The proof chain

Every figure is a citation, not text.

Each number in a Ward answer links to the rows it came from and the query that computed it, and it passed three checks before you saw it. Hover one below.

Try it

One answer, five figures, five chains.

A sample grocery dataset. The last figure was rejected once on the efficient-query check and rewritten, which is the part worth seeing.

Ward chat Sample grocery dataset
YouWhy is dairy margin down this week?
Ward

Dairy gross margin fell to this week, from the week before. Most of the drop is shrink at Store #37, which ran against a estate average. Receiving logged short at the dock across four deliveries, so the cause looks like receiving, not theft.

Hover or tap any number to open its proof chain.
21.4% Dairy gross margin, week 45
  1. 1Source rows
    POS export · synced into the lake · transactions ⋈ item_costs · 1,836,410 line items read
    storeskunet salescost
    3741022 Whole milk 1 gal$4.29$3.31
    5241870 Greek yogurt 32 oz$5.99$4.48
  2. 2Query
    from
    transactions, item_costs
    filter
    department = dairy · week 45
    group
    none
    measure
    (net sales − cost) ÷ net sales
  3. 3Checks
    • Valid data1.84M rows, no nulls in cost, week 45 complete
    • Efficient queryRead 1 of 52 week partitions
    • Answers the questionMargin for dairy, this week: matches the question
Returned only after all three passed.
24.0% Dairy gross margin, week 44
  1. 1Source rows
    POS export · synced into the lake · transactions ⋈ item_costs · 1,791,055 line items read
    storeskunet salescost
    3741022 Whole milk 1 gal$4.29$3.27
    1440115 Butter 1 lb$4.79$3.52
  2. 2Query
    from
    transactions, item_costs
    filter
    department = dairy · week 44
    group
    none
    measure
    (net sales − cost) ÷ net sales
  3. 3Checks
    • Valid data1.79M rows, week 44 complete
    • Efficient queryRead 1 of 52 week partitions
    • Answers the questionThe comparison week the answer names
Returned only after all three passed.
4.2% Shrink rate, Store #37, week 45
  1. 1Source rows
    POS export · synced into the lake · inventory_adjustments · 9,412 adjustment rows read
    storeskureasonunits
    3741022 Whole milk 1 galshrink−18
    3741870 Greek yogurt 32 ozshrink−11
  2. 2Query
    from
    inventory_adjustments
    filter
    store = 37 · department = dairy · week 45
    group
    store
    measure
    shrink units ÷ received units
  3. 3Checks
    • Valid dataAdjustment reasons all in the shrink code set
    • Efficient queryClustered on store, so one store read is one range
    • Answers the questionStore-level shrink, the figure the answer cites
Returned only after all three passed.
1.8% Shrink rate, estate average, week 45
  1. 1Source rows
    POS export · synced into the lake · inventory_adjustments · 402,977 adjustment rows read
    storeskureasonunits
    1440115 Butter 1 lbshrink−4
    5241022 Whole milk 1 galshrink−6
  2. 2Query
    from
    inventory_adjustments
    filter
    department = dairy · week 45
    group
    none
    measure
    shrink units ÷ received units
  3. 3Checks
    • Valid dataEvery store reported for week 45
    • Efficient queryRead 1 of 52 week partitions
    • Answers the questionSame measure as Store #37, so the two compare
Returned only after all three passed.
312 cases Cases short at receiving, Store #37, week 45
  1. 1Source rows
    SAP · read-only replica, queried in place · goods_receipts ⋈ purchase_orders · 1,204 receipt lines read
    POdeliveryorderedreceived
    918274Tue 06:1014096
    918390Thu 05:5512081
  2. 2Query
    from
    goods_receipts, purchase_orders
    filter
    plant = 37 · department = dairy · week 45
    group
    delivery
    measure
    ordered − received, in cases
  3. 3Checks
    Attempt 1 rejected on efficient query. It read every plant's receipts for the year. Ward rewrote it with plant and week filters and ran the checks again.
    • Valid dataEvery receipt joins to a PO
    • Efficient queryRewritten with plant and week filters
    • Answers the questionShortfall at the dock, which is what the answer claims
Returned only after all three passed.
The three checks

Passed before the answer returns, not flagged after.

Check 01

Valid data

The rows exist, the types match the measure, and the window is the one asked about.

Check 02

Efficient query

The query uses the partitions and indexes it should. A full scan where a filter would do is rejected.

Check 03

Answers the question

The result is checked against what was asked. A correct number for a different question does not pass.

A failed check loops back. You are not shown a figure with a warning next to it; the query is rewritten and the checks run again, and only a figure that passes all three is returned.

Why it holds

The model never does the math.

An AI analyst that pastes rows into a model and asks for a total is asking for invented numbers, and paying by the token for them. Ward splits the work.

The model

Reads the question, writes the queries and the sentence

It sees the question, the schema it needs and the computed results the asking user may see. Never your tables.

Ward’s data lake

Computes and checks every figure

Columnar, indexed, partitioned and clustered, so the math runs where it is exact and cheap.

Small context windows are also why checking is affordable. Verification is included in the 5% compute fee and adds little to the model bill. Pricing →

When the data is thin

A dead end comes back as the next question.

When the connected data cannot support an answer, Ward says so and suggests questions it can answer. No guess dressed as a figure.

Sample

“What did the Easter promo do at Store #52?”

I can't answer that from what is connected. Store #52 has no promo calendar rows before March, so there is no Easter window to measure. Try one of these:

Promo lift at Store #52 since MarchEaster promo lift across all storesWhich stores are missing promo calendar rows
Every week

Scheduled reports that say what changed.

Schedule a dashboard and Ward regenerates it on time and describes what moved since the last run, with the same proof chain on every figure. Ask a follow-up on the same data from there.

Questions

What people ask about the proof chain.

The record behind one figure in an answer: the source rows it came from, the query that computed it, and the three checks it passed. In Ward every figure in an answer is a citation rather than text, so hovering it opens that chain.

Valid data: the rows exist, the types match the measure and the window is the one asked about. Efficient query: the query uses the partitions and indexes it should, and a full scan where a filter would do is rejected. Answers the question: the result is checked against what was asked, so a correct number for a different question does not pass.

It loops back. The failed figure is not shown to you with a warning; the query is rewritten and the checks run again. Only a figure that passes all three is returned.

Ward says so and suggests questions the connected data can answer. A thin answer comes back as the next question, not a guess.

No. The model turns your question into queries and writes the sentence. Ward's data lake computes every figure, and the model receives the result. That is the main reason Ward's answers do not contain invented numbers.

No. Ward's query format is JSON. The proof chain shows its parts in plain terms: what it reads from, how it filters and groups, and what it measures.

Hover the first number Ward gives you.

Connect one read-only source. If a figure cannot show its work, you will see that too.

Read-only to start · your LLM keys · SOC 2 Type II underway · or book a call directly

Find out what your data has been hiding.

Tell us about your operation. We’ll show you the problems Ward catches, and the ones your current tools miss.

Step 1 of 3
What are your goals?
Step 2 of 3
About your operation
Step 3 of 3
Your contact info