Disaster Recovery at the Speed of AI: Susan Hunt Stevens of Tessi.ai — IAQ Radio Episode 782
IAQ Radio's Episode 782 (air date June 26, 2026) welcomes Susan Hunt Stevens, co-founder of Tessi.ai: an AI-first platform that, in the show's published words, generates universal damage assessments, certifies funding sources, and dispatches vetted workers — all before traditional systems send their first adjuster. Stevens is an award-winning founder building technology at the intersection of climate resilience, housing, and human behavior. Our curation prints the episode's published account at its true weight — a vendor's platform described by its founder on an industry show — and reads it against this site's own documentation discipline: an assessment, however fast, is a record somebody still has to stand behind.
Cluster: Moisture & Building Science. Creator score: 8/10 in our source registry. This page curates the creators’ own episode, embedded from their channel — we never download or re-upload. Our summary and key points are in our words, drawn from the episode’s own published description on iaqradio.com.
What this episode is
A founder interview about the paperwork end of disaster. In Episode 782 (air date June 26, 2026), IAQ Radio welcomes Susan Hunt Stevens, co-founder of Tessi.ai. The show’s published account is a single dense paragraph, and this curation prints its claims at exactly the weight of that paragraph: disasters move fast; Tessi’s AI-first platform generates universal damage assessments, certifies funding sources, and dispatches vetted workers — all before traditional systems send their first adjuster. Stevens herself is billed as an award-winning founder and CEO building technology at the intersection of climate resilience, housing, and human behavior, an expert in designing technology for behavior change, leveraging AI to make repairing homes after natural disasters faster, safer and smarter; in parallel she maintains a small personal climate investment and advisory platform (Spark Global) supporting board roles, entrepreneur-in-residence work, and select early-stage climate companies. The published description does not itemize interview segments, demonstrations, or results, and this curation invents none — the key points below are drawn from that account and from what the account’s own verbs claim, labelled as claims.
Key points (our summary of the published account)
- The bottleneck this platform names is real, and this site’s readers live in it. After a flood or hurricane, the first scarce resource is not a contractor — it is an assessment: the documented statement of what is damaged, which is what funding, insurance, and dispatch all wait on. A mold library notices the sequence because mold runs on the same clock in reverse: every day a wet home waits for its first adjuster is a day inside the 48-hour drying window this site’s NIOSH summary prints. A platform whose entire pitch is collapsing the wait before the first assessment is aiming at the exact delay that turns water damage into a mold case.
- Three functions, three different kinds of claim. “Generates universal damage assessments” is a documentation claim; “certifies funding sources” is a verification claim; “dispatches vetted workers” is a marketplace claim. The published account asserts all three in one sentence, as a founder’s account of her own platform on an industry program — which is how this page carries them: as the company’s stated design, not as audited performance. No figures — assessments completed, funding certified, workers dispatched, timelines measured — appear in the published description, and none are supplied here.
- Why it is curated anyway: the documentation standard is the argument. This site’s whole method is that a record beats a recollection. An AI-generated assessment raises the same question in a new costume as a member’s undated review: what is it evidence of, and who stands behind it? The interesting hour here is a founder whose background is behavior-change technology describing how a machine-produced assessment gets trusted by the humans — homeowners, funders, workers — who have to act on it. That trust question, not the software, is the episode’s transferable content for this shelf.
Claims discipline: the statements above are the show’s published account of its guest and her platform, presented as reported. Platform capabilities are the company’s claims, printed as claims; the show is industry-sponsored, stated in the creator card below. This site makes no finding about Tessi.ai’s performance, any assessment it has produced, or any facility or claim mentioned in the episode.
Why it is on this shelf
Video Watch mostly curates practitioners explaining their craft. This episode is the shelf’s one look at the layer above the craft: the assessment-and-funding machinery that decides, after every major storm, which homes get dried in 48 hours and which homes become the next decade’s mold files. Stevens arrives from outside the restoration trade — climate resilience, housing, behavior change — which is precisely why the conversation belongs next to the trade’s own: the field’s insiders (Spurgeon, Cole, Shaughnessy) define what a good assessment is; this episode is about whether one can be produced at machine speed and still be believed.
Creator card — IAQ Radio
- Who: IAQ Radio — “Radio Joe” Hughes and Cliff “Z-Man” Zlotnik, industry veterans whose live Friday show is, by its own billing, the world’s longest-running continuous live technical podcast; Episode 782 aired June 26, 2026, in the run-up to the show’s 20th anniversary (Ep. 784). Registry score: 8/10 — unmatched depth and guest caliber; practitioners who let credentialed experts and working inspectors talk, with a checkable record. Full profile: IAQ Radio in Voices.
- What they do: Weekly long-form interviews and working sessions with IAQ researchers, industrial hygienists, laboratory scientists, restoration practitioners — and, as here, the founders building platforms around the trade. Past curations: Episode 497 · Episode 573 · Episode 781 · Episode 783 (Cole, Part 1) · Episode 784 (20th Anniversary) · Episode 785 (Cole, Part 2) · Episode 786 · Episode 787
- Commercial flags, stated factually: The show is industry-sponsored, so vendor voices appear alongside the scientists. This episode’s guest is the co-founder of the platform being discussed: the episode is, in part, a vendor presenting her own product on an industry program, and this curation carries it labelled as exactly that.
- Their doors: IAQ Radio — homepage (Episode 782, air date 6-26-26)
Tools for this job: Measure — moisture is the whole game
Pinless moisture meter
What it is: A handheld meter that reads moisture inside drywall, wood, and flooring through the surface — no pinholes, no damage. The tool that finds the wet wall behind the paint.
Why it is listed: Moisture readings are the hardest evidence to argue with. A dated photo of a meter reading on a wall turns a landlord dispute from opinion into documentation.
Its limits: Readings are relative and material-dependent — use them comparatively (wet spot vs. known-dry spot) and record both. A meter finds moisture, not mold.
Thermo-hygrometer / humidity data logger
What it is: A room thermometer-hygrometer; the logging kind records humidity over days and weeks instead of showing a single moment.
Why it is listed: A log beats a snapshot. Two weeks of humidity history above 60% is a document; a glance at a dial is an anecdote.
Its limits: Cheap sensors drift — sanity-check a new unit against a second one. Placement matters: measure where people live and sleep, not next to a shower door.
Some links in this strip are affiliate links and cost you nothing extra if you buy through one. As an Amazon Associate I earn from qualifying purchases. Affiliate money never buys scores, coverage, or silence — see our affiliate disclosure.
If this happened to you
Waited weeks after a storm for a first assessment while a wet home grew worse — or got a fast assessment you believe was wrong? File a report with the dates: the storm, the request, the assessment, the work. The timeline is the record, whichever system produced it.