The Intermediary –- May 2026 - Flipbook - Page 94
T E C H N O L O GY
Opinion
Augmentation is
the future
W
hat is the
difference
between
automation
and
augmentation
in lending? While automation seeks
to replace human decision-making
with algorithms, augmentation uses
technology to remove administrative
friction, allowing human experts to
focus on complex risk assessment.
In the broader world of financial
services, innovation is frequently
used as a synonym for automation.
The industry’s goal is usually to build
a conveyor belt: clean data goes in,
a binary decision comes out, and a
human never has to touch the file.
If you’re lending on a three-bed semi
to a teacher with a fixed salary, that
conveyor belt works. But in specialist
lending, the conveyor belt breaks
every single day.
Whether it’s a multi-layered buyto-let (BTL) structure, an auction
bridge, or a ground-up development,
we deal in non-standard. In our world,
complexity isn’t a hurdle to be cleared;
it’s the defining reality of the market.
The mistake many make is thinking
that more technology should lead to
less human involvement. I believe the
opposite is true.
Why ‘black box’ fails
Traditional PropTech tries to force
every deal into a rigid ‘Yes/No’
decision tree. But a computer doesn’t
intuitively understand the nuance of
a borrower with a complex corporate
structure or the specific potential of a
‘grey-belt’ planning site.
When you try to automate the
judgment in these cases, you end up
with one of two outcomes: a ‘no’ to a
perfectly viable deal because it didn’t
fit a pre-defined box, or a ‘yes’ to a
risk that a human expert would have
flagged in seconds.
In specialist lending, human
intuition isn’t an inefficiency; it’s the
92
The Intermediary | May 2026
core of effective risk management.
The consequences of geing this
wrong are more significant than they
might appear. A false ‘no’ doesn’t just
lose a deal, it damages the broker
relationship that brought it to you,
signals to the intermediary market
that your appetite is narrower than it
is, and ultimately costs you pipeline
you’ll never see.
In a market built on relationships
and reputation, that’s an expensive
failure mode. The deals that
automated systems decline are oen
precisely the deals that specialist
lenders exist to do.
Empowering underwriters
At LendInvest, our approach isn’t
about automating the underwriting
decision; it’s about automating the
administrative burden that slows
it down.
The reality of specialist lending is
that high-level experts oen spend
a lot of their day on ‘data friction’
chasing anti-money laundering (AML)
trails, manual portfolio stress testing,
and verifying messy asset valuations.
That isn’t underwriting; it’s data entry
and a poor use of talent.
We use technology to synthesise
those disparate data points into
a single, structured overview.
This doesn’t replace underwriting
intuition; it complements it.
In practice, this means an
underwriter arrives at the decision
point with everything they need
already assembled, valuations crossreferenced, AML documentation
verified, portfolio exposure calculated
and relevant precedents surfaced from
past cases.
What would previously have taken
hours of administrative work happens
before they open the file. The time
saved isn’t marginal. It’s the difference
between an underwriter making two
decisions a day and making six.
Critically, those decisions are beer
informed because the technology has
JASMIN ALLOTT
is director of technology,
product and data at LendInvest
already done the job of making the
complexity legible.
Human-AI collaboration
I don’t see an immediate future where
AI approves a complex development
loan in a vacuum. That isn’t a world
that, at this point, manages risk well.
Instead, I see a relationship.
Technology with the perfect
memory instantly summarising
hundreds of pages of reports or
identifying paerns from past cases.
The human providing contextual
judgement, understanding the local
market sentiment, the developer’s
track record, and the instinct that
comes from years in the trenches.
The verdict
The most effective way to scale
specialist lending in 2026 is not
through total loan automation, but
through tech-augmentation that
removes administrative friction while
keeping the human expert in control.
Simplification in our market
doesn’t mean the deals get easier. The
structures will stay intricate, and the
assets will stay unique. However, we
can make the process of geing those
deals done friction-free.
The future of specialist lending isn’t
a hands-off machine; it’s a high-tech,
high-touch environment where the
tech takes the complexity tax away,
allowing the specialists to do what
they do best.
We don’t need tech to make the
decisions for us; we need it to give
us the clarity to make beer and
faster ones. ●
