What it does for the person holding it
Homeowners do not decide who to believe by weighing arguments. They decide by whether they can follow what they were told. An inspector who can show the measurements, name the method, and say plainly where their opinion stops is believed. One who insists is not.
Your standard
Their trust
You did the same walk you always do. The difference is that now it is written down in a form that survives the next person who reads it.
The report carries your firm, your licence number and your professional opinion. The software formats and computes — the judgment on the page stays yours, and the limits it states are what make the rest of it credible.
Where the line is
Texas draws two hard lines through this work — one at the practice of engineering, one at the adjusting of insurance claims. Crossing either is not a technicality. This document is shaped so that what you hand a homeowner is measurement and record, and so the calls that are not yours to make are routed to the person licensed to make them.
A TREC inspector must “render a written opinion as to the performance of the foundation” — 22 TAC § 535.228(a). The Standards of Practice prescribe no method for reaching it: elevation surveys, deflection and tilt appear nowhere in the rule. The one opinion the state requires of you is the one it leaves you to justify alone. This gives it a measured, repeatable basis you can put in front of anybody.
TBPELS Policy Advisory Opinion 43 holds that a visual performance opinion “does not engage in the practice of engineering” — but that “analyzing the cause of a condition, recommendations for repair… would be considered the practice of engineering”, under Tex. Occ. Code Ch. 1001. The document records measured condition and hands cause and repair to a licensed engineer, in its own text, on the page — where a reader can see the handoff.
“A contractor may not act as a public adjuster or advertise to adjust claims” — Tex. Ins. Code § 4102.163(a). Violation is a Class B misdemeanor (§ 4102.206), and the customer’s contract becomes voidable with nothing owed for work already performed (§ 4102.207). The Texas Supreme Court upheld both provisions in TDI v. Stonewater Roofing on 7 June 2024. A measured record of what a structure is doing is documentation — it is what you produce instead of an opinion about the claim.
Under the Residential Construction Liability Act a claimant must produce “any expert reports, photographs, and video or audio recordings” that would be discoverable — Tex. Prop. Code § 27.004(a). Whatever you wrote becomes part of that file whether you intended it to or not. A document carrying its instrument, its calibration date, its raw readings and its arithmetic survives that journey. A one-line note on a work order does not.
This is not a state-approved or state-certified document, and no Texas law requires it. The ASCE Texas Section guidelines are a voluntary professional-society methodology; the document says of itself that its guidelines “are not intended to be Standards”. The terms Level A and Level B come from that guideline and appear in no Texas statute or rule. Anyone selling a “state-certified foundation report” is selling something that does not exist. What this is: the methodology Texas engineers actually use for residential foundations, applied the same way every time, with every number traceable to the reading it came from.
What a proper inspection is worth
A homeowner deciding between three contractors is not comparing methodologies. They are deciding who they believe. The measured, readable, self-limiting document is not a nicer deliverable — it is the reason they pick you, and the reason they call you back in six months instead of somebody else.
A colour-coded map of their own floor plan with a number at every point. Understanding is what converts into agreement — nobody signs for work they cannot picture.
The report names its method, logs the instrument, and states what it does not conclude. People trust a professional who marks their own limits more than one who has an answer for everything.
The survey that finds the low corner is the one that does not come back to haunt you. Being wrong about "minor" is how a customer relationship ends.
Every property becomes a baseline. A monitoring visit six months on is a reason to return, a second measurement that means something, and a customer who already trusts you.
The part nobody talks about
You finished the inspection at four. You finished the report at nine.
The inspection isn't the hard part. You've walked a thousand slabs — you knew what was happening before you got the level out. The hard part is turning what you saw into documentation that holds up when the homeowner, the adjuster, or the next contractor reads it.
How it works
Photos, elevation readings, notes on your phone. Nothing new to learn, no new gear to buy.
Point by point, with your observations. Calculations run live against residential thresholds as you go.
Findings, severity, and recommendations — formatted and ready to send to the homeowner.
Open it yourself
Not a video, not a guided tour. The real application with a complete case already in it — an initial evaluation, a six-month monitoring visit, sample hydrostatic and CCTV evidence attached to specific sewer segments, and § 7.0 comparison computed between the two. Click through every section. Change a reading and watch the numbers move.
The demonstration uses its own storage key, so nothing you do in it touches a real project. Everything stays in your browser; nothing is sent anywhere.
The same house, four different surveys
One floor plan. Four sets of readings. The settlement field is interpolated from your points every time, so the picture has to change when the measurements change — there is no library of pre-drawn severity graphics to pick from.
Inverse-distance weighting across every placed point, resampled on a 12 px grid.
0.5 in. / 1.0 in. / 1.5 in. — the firm's stated convention, named as its own in the report.
Green is within tolerance, red is where the slab is low. A homeowner needs no training to see the shape.
Verified, not asserted
Before publishing a claim about this tool, we generated thirty complete reports across four settlement patterns, three severity bands, 10 to 18 measurement points, and five pipe material and era combinations. Here is what came back.
Figures from our own test run of the shipped engine, August 2026. A report that documents an absence of movement is the one that closes a real-estate transaction — so the within-tolerance case was tested as hard as the critical one.
Inside the report
Not a checklist with a photo grid. A structured record — every measurement traceable to a point code, every threshold traceable to a published standard, and a plain-language summary for the homeowner bound into the back.
Four pages at full size. Click any one to read it.
Real generated output from a real inspection. The address, the owner and the inspector have been permanently removed — from the image pixels and from the PDF text layer both, using true redaction rather than drawn boxes. Everything else is exactly as the engine produced it.
What gets computed
Figures shown are from the sample report above.
§ 3.5.2 — the step most tools skip
Floor plans upload at whatever resolution they happen to be. Until the tool knows how many pixels make a foot, every derived length — sewer segment runs in § 3.6, the deflection span in § 5.2, and every slope computed from them — rests on a guess. Most software never asks. This one refuses to pretend.
The longer the baseline, the smaller the error it carries into everything downstream.
One tape measurement, taken once per property.
Pixels per foot, and the building envelope that implies. If the envelope does not match the footprint, the points were wrong.
§ 3.5, § 3.6 and § 5.1 all print "assumed scale" rather than presenting a guess as a measurement.
On the bundled demonstration, calibrating from the 20 px/ft default to the measured 13.44 px/ft moves the deflection ratio from L/281 to L/419 — across a classification boundary.
That is the difference between a report that says "repair recommended" and one that does not. It is one tape measurement.
Three kinds of visit
A single inspection cannot tell a seasonal drift from active movement — and that distinction is what an adjuster, an engineer and a judge actually want to know. Every project holds an unbounded series of visits, and the point grid carries forward so the same labels are re-measured every time.
Initial
The baselineFull survey, plan placed, scale calibrated, sewer traced. Everything downstream is measured against this.
Monitoring
Six months laterPoint labels and positions copy forward; readings clear for re-measurement. § 7.0 computes interval, per-point delta and rate in inches per month, with the datum shift removed in its own column.
Post-repair
Did the work hold?Near-zero movement documents stabilisation. Continued movement documents remediation failure — with numbers, against the original baseline. It is called a Survey, never a Certification.
Points introduced at a later visit are shown but excluded from rate calculation, because a point with no baseline has no rate. The report says so rather than quietly averaging it in.
§ 6.7 — the handoff
A licensed P.E. asked to issue a sealed opinion should not have to re-survey the property or retype your table. One button emits the underlying data as CSV, wrapped in a header that names the property, the firm and the inspector — and states outright that no engineering opinion is asserted in the file.
Every point from every visit — visit id, date, type, label, plan position, elevation, notes.
Every segment from every visit — run, as-designed and current slope, joint-stress index, classification, evidence count.
This is what keeps the firm inside its scope of practice: the measurement leaves, the judgment stays with the person licensed to make it.
Why the numbers hold up
A template fills in blanks. This computes, and then it shows its work — because the first thing a reviewing engineer does is try to break your arithmetic.
ASCE Texas Section § 5.4 defines the deflection ratio as the maximum deviation of a third point from a straight line between two points, divided by the distance between them. Every pair of plan-located points is searched as a candidate chord and the governing — worst — triple is the one reported. The report names it: “0.54 in. deviation at M-08 from the 16.2 ft chord M-03 to M-11.” Not the pair that happened to look bad on site.
A house can lean without bending. Tilt is rigid-body rotation; deflection is bending, and only bending cracks brick. Reporting them as one number is the most common error in foundation documentation. Here they are two numbers — tilt taken from a least-squares best-fit plane through every plan-located point, printed on its own line beside the deflection ratio.
Where the tool makes a judgment the standard does not dictate — the chord-acceptance tolerance, the severity banding — the report says so in its own text: an intermediate point is accepted when it falls within 10 percent of the chord length in plan, “which is this tool’s search convention and not a figure from the guidelines.” Of the severity bands it says they “do not establish distress, and they assert nothing about cause.” A reviewer can accept that, adjust it, or reject it. What they cannot do is be misled by it.
A relative survey carries no absolute benchmark, so each visit is reduced to deviations from its own reference point. When two visits are compared, the change of datum is subtracted from every figure — reported as a separate column — so what survives is the change in shape of the structure, not the arbitrary zero you started from.
As-designed slope, the current slope computed from measured settlement at each end, and the IPC minimum, on one axis. Vertical scale is exaggerated for readability and the exaggeration factor is printed on every profile — so nobody mistakes the drawing for the magnitude.
The joint analysis ranks which sewer segment to test first — from measured differential settlement, the joint tolerance of the pipe material and the installation era. It tells a crew where to put the camera. It is not a leak test and does not replace one: the report is a reason to test, never a substitute for testing.
Drain slope minimums come from IPC Table 704.1 and IRC Table P3005.3 — ¼ in./ft for 2½ in. and smaller, ⅛ in./ft for 3–6 in. Evaluation method comes from the ASCE Texas Section Guidelines, Version 3, adopted 1 April 2022 — cited by edition in the report’s own reference list. Nothing in the report is a number this firm invented.
Where the evidence comes from
Published survey data
Site geology is recorded in § 2.3 and printed in the report: soil map unit and dominant component, shrink–swell potential, plasticity index, linear extensibility and drainage class. On the sample property that reads Houston Black clay, map unit HuB, shrink–swell Very High, PI 40–60, LEP 9.5% — so a reviewer can see what the analysis assumed about the ground.
Cited for regional context. The report says outright that this is not a parcel-specific geotechnical determination — that requires a site-specific study by qualified professionals.
Instrument provenance
Every report logs the instrument, its serial number, its calibration date, its stated accuracy (±0.05 in.), the unit and the reference mode. A reading with no instrument history behind it is an anecdote. A reading with a serial and a calibration date is a measurement.
This is the section a reviewing engineer or an opposing expert turns to first. It is filled in every time, because the tool will not let it be skipped.
Nothing is invented. Missing readings, missing license details and missing calibration dates are flagged for you to supply — never filled in by the tool. That single rule is what makes the rest of the document worth anything.
Part II · bound into every report
The technical sections are for the engineer, the adjuster and the file. Part II is written for the person standing in the kitchen deciding whether to believe you — plain language, one map, and a numbered list of what happens next.
This is the part that gets forwarded — to a spouse, an insurer, a second contractor. It carries your name on it.
What it does for you
Most contractors already know what they found before they leave the property. What they don't always have is a defensible next step, written down, in the right order.
Hydrostatic pressure test and CCTV inspection by a licensed plumber; comparative re-survey at a defined interval to establish whether movement is active and at what rate; soil moisture sampling at the perimeter if an engineer determines it material. Each tied to the specific open question it answers.
The joint-stress index ranks the sewer run so the plumber's scope goes where the measured differential is worst — instead of scoping the whole line and hoping.
Causation and remedial scope are engineering judgments reserved to a licensed P.E. under Texas Occupations Code Ch. 1001. The report states that in its own scope section — which protects you as much as it protects the homeowner.
Verification testing before remedial work both preserves the evidentiary record and establishes whether an active leak is a contributing factor. Do it in the other order and the question can never be answered.
Built for Texas
Expansive soils swell and shrink with every wet season and drought. That movement shows up as stair-step cracks in brick, doors that stop latching, and floors that read an inch out of level across 20 feet.
None of these proves failure on its own. Together with elevation readings, they build a case — and that case is what the report is for.
Scope, stated plainly
What the report does
What it does not do
Determination of cause is an engineering judgment reserved to a licensed Professional Engineer under Texas Occupations Code Chapter 1001. Drawing that line clearly is what makes the measured record usable by the engineer who reviews it.
Platform tier
The agent writes the report. The platform is the working instrument behind it — for crews running monitoring intervals and building an evidentiary record over time.
Runs where the work is
A single file that opens in the browser you already have. Your data stays in local storage on your own device, and the PDF engine is bundled — so a crawlspace with no bars is not a problem, and no client's address is ever sitting on someone else's server.
What you are actually buying
Not a faster way to type. The difference between "the contractor said the slab is moving" and a measured record, on a published standard, that an engineer can pick up and review.
Without it
With it
One report that does not hold up costs more than a year of this. Not in fees — in the job you do not get called back for, and the homeowner who tells three neighbours. The documentation is the part of the work that outlives the visit.
Read the rest
The system overview, the field workflow, the regulatory reasoning and the full release history are published. So is the source of every page you have just read.
What we found in our own work
In August a verification pass recomputed the calculation engine from scratch, independently of the code that produced it. It found five things wrong. We fixed them and wrote down what they were, because a measurement tool that has never been audited is just a tool nobody has checked yet.
The engine computed elevation range over a span, called it deflection, and classified it against L/360 and L/240 — which are bending limits. A slab that was flat but out of level read as bent. Deflection is now the departure from a least-squares plane; tilt is reported separately as that plane's gradient.
And duplicated as a bare 0.05 in two other places. Every segment run and every span depended on it. § 3.5.2 calibration was added, and until it runs the report says its lengths are assumptions.
Screen, PDF and CSV export could each show a different number. There is one function now.
The flagship multi-visit feature existed on screen only. It prints now — and cross-visit datum shifts are normalised before differencing, so a change of survey reference can no longer masquerade as structural movement.
Peak-to-trough measurement was replaced with the ASCE § 5.4 chord construction. Threshold tiers that do not appear in the source guidelines were removed rather than kept for convenience.
Read the whole history in the changelog. A vendor who will not tell you what they got wrong has not stopped getting things wrong.