Detection and classification
Finds candidate items in photographs and suggests a material category, showing uncertainty for ambiguous cases.
Material recovery planning
Salvora helps deconstruction teams identify and organise potentially reusable building materials. AI turns site photographs and survey documents into a proposed inventory that surveyors can verify, prioritise, and export for recovery planning.
Suggestions are unconfirmed until a surveyor verifies them.
Why Salvora exists
Every building is a store of materials someone already made, carried and installed. Too often it leaves site as rubble because no one had time to see what was worth saving. We want the reuse potential found before the first wall comes down.
Our purpose
Help deconstruction teams find, verify and plan the recovery of reusable building materials.
Why it matters
Construction and demolition produce some of the largest waste streams in the world. Recovering even part of what buildings contain saves materials, money and carbon, but only if it is identified in time.
The stakes
Construction and demolition debris generated in the United States in 2018.
Source: US EPAShare of total EU waste generation from construction in 2022.
Source: Statista, citing EurostatThe European Commission’s figure for construction’s share of EU waste, confirmed by fact-checkers.
Source: EUFactcheckFigures are published estimates from the sources linked; they describe the wider problem, not results from this product.
What changes
Today
With Salvora
The problem
Buildings can contain doors, flooring, fixtures and components worth recovering. Finding them means walking every floor, then reconciling photographs, plans, schedules and notes into one list. Opportunities are missed, and inventories end up with uncertain quantities, duplicates and gaps in condition information.
The solution
Salvora detects candidate items in site photographs, suggests material categories with uncertainty, groups photos that may show the same item, pulls references from survey documents, and helps rank confirmed items for recovery.
Finds candidate items in photographs and suggests a material category, showing uncertainty for ambiguous cases.
Groups photographs that may show the same physical item, so counts are not inflated.
Extracts item references from schedules and surveys, and ranks confirmed items by condition, access and reuse demand.
How it works
Add room photographs, plans and existing survey records for the project.
Salvora suggests candidate items, linked to the images and documents they came from.
Surveyors confirm categories, quantities, dimensions and condition, and add access and inspection notes.
Confirmed items are prioritised and exported as a recovery inventory and inspection plan.
Priorities are calculated only from surveyor-confirmed inputs. Unconfirmed suggestions never reach the exported inventory.
Who it’s for
Plan recovery before strip-out, with quantities you can price.
More recovered materialSee what a project could supply before committing to buy.
Better sourcing decisionsUnderstand reuse potential early in a refurbishment.
Reuse built into the planProduce pre-demolition inventories faster and more consistently.
Faster, consistent surveysNamed organisation types are prospective customers. No affiliation or endorsement is implied.
Honest limits
Salvora helps teams find and organise candidate items. Whether a component is structurally suitable, free of hazardous substances or acceptable under regulations is for qualified people to establish.
AI, used carefully
A recovery inventory has to stand up on site. Salvora speeds up finding and organising items, and leaves verification with surveyors.
What the AI does
What it never does
Every candidate item is confirmed, dismissed or escalated by surveyors, never by the model.
Each flag links to the site photographs, plans, schedules and survey records behind it, so a reviewer can check the reasoning before acting.
Scores come with a confidence level and the data gaps that lower it, so weak signals are not mistaken for facts.
Permissions are set per role: who can view, comment, assign, decide, escalate or export.
Assignments, notes and decisions are time-stamped with the person and reason, and can be exported.
Customer data is isolated per tenant, encrypted at rest and in transit, and never used to train models for other customers.
What we believe
Nothing enters the inventory until a surveyor confirms it.
Each entry links to the photos and documents behind it.
Ambiguous materials are marked as ambiguous.
Structure and hazards are for qualified people.
Interior items first, done well.
Duplicate observations are resolved, not multiplied.
A note from the founders
Ikenna has walked buildings days before strip-out, photographing doors, panels and fixtures that would have been worth saving, knowing there was no time to turn those photos into a usable list.
Toyin works with organisations trying to reuse more materials and sees how often recovery fails for lack of reliable quantities. Mustapha builds vision models and knows that a proposal is only useful if people can check it quickly.
Salvora is designed to do the tedious part, finding and grouping candidate items, so surveyors can spend their time verifying, prioritising and planning recovery.
Ikenna Obiora, Founder/CEO · Toyin Adebanjo, Co-founder · Mustapha Aliyu, CTO
Where we’re going
From interior items to portfolio-wide recovery planning.
Now
An interactive prototype on fictional survey data, and conversations with contractors.
Next
Photo-based suggestions, duplicate review, confirmation, inspection status and export for doors, cabinets and floor panels.
Then
Fixtures, finishes and further categories as models are validated.
Later
Recovery reporting across many buildings, with integrations for partners.
Pricing
Project-based assessment fee, or a subscription for contractors handling multiple sites.
For a single assessment.
$1,800/ project
For teams running several sites.
$950/ month
For developers and large contractors.
Custom
Illustrative pricing for a proposed product. Final pricing is agreed with early-access customers and excludes taxes.
Proposed production design
How Salvora could run in production, from site photographs to an evidence-linked recovery inventory.
Site photographs, plans and survey documents per project.
Detection, segmentation and material classification models.
Item references and specifications from supported survey documents.
Inventory, review decisions, inspection status and evidence links.
Recovery summaries that cite the images and documents behind them.
Access logging and encryption per customer.
Proposed, not deployed. No AWS services are connected to this website or the prototype, which run in your browser on fictional data.
The team
A proposed founding team for this concept.
Founder/CEO
Pre-demolition surveyor with experience on commercial refurbishment projects.
Co-founder
Circular economy consultant focused on material reuse in construction.
CTO
Computer vision engineer working on detection and segmentation.
Nonstructural commercial-interior items such as doors, cabinets and raised-floor panels. Structural assessment and a resale marketplace are deliberately out of scope.
No. Hazardous-substance questions need qualified surveys and testing. Salvora can record that an inspection is required and track its status.
Possible duplicate photos are grouped for review, and only surveyor-confirmed quantities go into the exported inventory.
Not yet. Salvora is a proposed startup concept. The team is fictional, pricing is illustrative and the prototype runs on fictional data in your browser.
No. Salvora flags, ranks and explains. Surveyors review the evidence and record every decision with a reason.
In production, data would be encrypted, isolated per customer, access-controlled by role and never used to train models for other customers. Data-sharing terms are agreed before any pilot.
Early access
We are looking for a small number of early partners to shape Salvora. A pilot starts with a sample of your historical data and a fixed review period, so you can judge the results before committing.