Human Touch™
The tool is innovative.
The experience is magical.
Data quality software that can read and process data as easily as we do.
At the core lies our most advanced Human Touch™ technology.
This powerful software features large, highly capable intelligence models specifically tailored to read and process data as efficiently as humans do.
The Human Touch™ logic intelligently interprets the full address, making sense of the information in the same intuitive way a human would. With the inclusion of intelligent location awareness, fuzzy logic, phonetic matching, bordering localities, and many other expansive methodologies, we continue to push the limits for what it means to have the most accurate data possible.
Human-like intuition
Incorrect or poorly entered data. Let's look at an example of an address that users have entered manually. Here you can see an address that has been typed in with no care for format and missing details. Instead of just accepting this address, which would likely cause problems downstream, we're going to pass it through Human Touch™ technology.
And like magic, you can see that the address has not only been formatted correctly, but it has also had misspellings repaired and missing information added.
Let's take a closer look at how Human Touch™ tackles the most common scenarios, where an address looks right at first glance, but quietly hides errors and inconsistencies.
Small details making
big differences.
- Recognising that "North" could be part of the street name ("Old North Rd") or a directional suffix ("Epping North").
- Knowing how suburbs, postcodes, and states interact, so even if one field is missing or wrong, the system can infer the likely correct location.
- It's like giving the software local street smarts.
- Match "Minto Avnue" to "Minto Avenue."
- Recognise "Strret" as "Street." Work with near-miss errors, typos, and formatting inconsistencies instead of rejecting them outright.
- "Macauley Rd" matching "McAuley Rd."
- "Smithson" matching "Smythson."
This is particularly useful for names derived from many cultural origins where spelling variations are common.
Geographic Intelligence
- Checking if the given suburb sits right on the edge of another and correcting it accordingly.
- Fixing vanity suburbs (when someone writes a "prestige" suburb instead of the official one).
Purposely substituting a suburb or town for a more prestigious one nearby is more common than you think. Human Touch™ catches these errors to prevent misdirected deliveries, tax zone issues, and misaligned service areas.
- 34–36 Darlinghurst Rd (a property that spans multiple street numbers), or 21A–21C (multiple units within the same number).
Human Touch™ recognises and interprets these ranges and alphanumeric suffixes correctly. Instead of flagging them as invalid or splitting them incorrectly, it understands they're part of how properties are officially designated.
- In rural or undeveloped land, "Lot 5" could be the legal land parcel identifier.
- In urban areas, "Lot 5" might sit inside a unit block, needing different treatment.
Human Touch™ applies this context, so the same label is understood differently depending on its setting. That prevents errors like misplacing rural addresses into city-style formats.
The beauty is in the details.
Standardisation on Invalid Address Entries.
While we can't make all addresses valid, sometimes it's simply not possible, we don't walk away from the challenge. With Seltaris, we take a different approach. We aim to make all your data as clean, consistent and correctly formatted as possible, even those addresses that can't be fully validated. Most providers are happy to boast about an 85% validation rate and quietly ignore the remaining 15%. But when you're dealing with 100,000 addresses, that 15% isn't a small problem, it's 15,000 addresses that would otherwise need time-consuming manual effort to clean up and correct.
That's where the real headaches begin, leading to delays, extra costs and frustrated teams. We go the extra mile by not just validating what we can, but also working hard to improve the quality of the addresses that fall outside standard validation. We clean them up, correct obvious issues, validate and repair what we can and standardise them to bring them in line with your verified data. The result? Even your unvalidated addresses are in the best possible shape.
Individual field corrections.
While others might reject an address as unfixable if key parts are missing or invalid, we take a different approach. With Seltaris, even if all you have is a suburb and the rest of the address is incomplete or incorrect, we still validate and correct what's there. We process each individual field, clean up obvious issues, and standardise the data to bring it in line with your verified records. The result? You get the highest quality address possible, even if it's not yet deliverable, putting you in the best position for future updates and minimising costly manual cleanups down the track.Built from the ground up with only the finest datasets.
Designed to source the best datasets available from multiple suppliers in one system to give you the very best, utmost accurate results.
Quality datasets mean more accurate results
Premium dataset options
While most solutions rely solely on the standard government datasets available in Australia and New Zealand, Seltaris goes further.
Australian Postal Data
99% accuracy
GBG OptiSource System™ includes an expansive semantic index that organises and surfaces information from a variety of trusted data sources, including the Australian PAF, renowned for its 99% accuracy in postal data.
When a file request is made, Seltaris processes the data through its Human Touch™ algorithms to verify validity. It then searches its semantic index and feeds the results to Human Touch™ generative models, delivering optimal context to accurately format, cleanse, repair, verify, and standardise addresses.
This entire process happens in milliseconds, giving you the fastest and most precise results, every time.
Explore the core features
of Seltaris
Slide 1 of 5