While
much
of
the
AI
development
in
legal
tech
focuses
on
the
practice
of
law
—
research,
drafting,
document
review
and
the
like
—
Centerbase,
the
practice
management
platform
for
midsized
law
firms,
is
releasing
a
new
AI
feature
that
focuses
on
the
business
of
law.
Today,
the
company
is
announcing
the
limited
release
of
Centerbase
IQ,
an
AI-powered
natural
language
chat
tool
embedded
directly
within
the
Centerbase
platform
that
enables
managing
partners
and
firm
administrators
to
ask
questions
about
firm
performance
in
plain
language
and
get
quick,
visual,
citation-backed
answers
drawn
from
the
billing,
financial,
matter,
and
productivity
data
already
housed
in
Centerbase.
The
company
is
debuting
the
product
this
week
at
the
2026
Association
of
Legal
Administrators
Annual
Conference
&
Expo
in
National
Harbor,
Md.
For
many
firm
leaders,
getting
answers
about
firm
performance
still
means
submitting
a
report
request,
waiting
for
it
to
be
built,
and
then
receiving
information
that
may
already
be
outdated.
Centerbase
IQ
is
designed
to
eliminate
that
cycle.
“Managing
partners
should
not
need
a
help
ticket
to
understand
how
their
firm
is
performing,”
said
Michael
Dunn,
CEO
of
Centerbase,
in
a
statement.
In
a
pre-release
demonstration
last
week,
Rob
Joyner,
senior
vice
president
of
business
development,
framed
the
launch
in
the
context
of
Centerbase’s
broader
evolution.
The
company’s
first
act,
he
said,
was
building
the
core
system
of
record
—
billing,
accounting,
and
practice
management.
Its
second
act,
which
I
wrote
about
last
month,
has
been
connecting
AI-powered
tools
such
as
NetDocuments’
ndMAX
into
Centerbase
workflows
to
create
what
Joyner
calls
a
“fluid
data
layer”
for
midsized
firms.
Centerbase
IQ
represents
the
third
act,
putting
an
intelligence
layer
on
top
of
that
data
so
firms
can
make
sense
of
it
through
natural
conversation
rather
than
static
reports.
“A
lot
of
the
agentic
AI
that
you’re
talking
about
is
really
focused
on
the
practice
of
law,”
Joyner
told
me.
“Centerbase
IQ
is
agentic
AI
built
around
the
business
of
law.”
How
It
Works
During
the
live
demo,
Scott
Cormier,
Centerbase’s
chief
product
officer,
showed
me
how
the
interface
works.
Users
can
either
type
or
speak
questions
in
natural
language,
or
start
from
pre-configured
“kickstarter”
questions
organized
by
categories
such
as
financial
health,
matter
status,
intake
pipeline,
billing
methods,
and
more.
The
system
responds
with
a
combination
of
narrative
text,
embedded
tables,
and
charts
—
including
line,
bar,
and
pie
charts
—
along
with
citations
showing
the
source
records
behind
each
answer.
A
right-side
panel
displays
the
full
set
of
data
sources
used
for
any
given
response,
and
users
can
export
tables
and
visuals
or
continue
asking
follow-up
questions
in
a
threaded
conversation.
Cormier
demonstrated
queries
ranging
from
broad
financial
overviews
to
specific
operational
questions,
such
as
identifying
trends
in
flat-fee
arrangements
or
determining
which
timekeepers
logged
the
most
hours
in
a
given
month.
Each
response
included
proactive
insights
—
observations
and
recommendations
generated
alongside
the
raw
data.
The
tool
draws
on
more
than
20
data
entities
within
Centerbase,
spanning
matters,
clients,
attorneys,
billing
entries,
invoices,
payments,
documents,
calendar
events,
and
more.
Cormier
credited
Joyner
for
pushing
for
the
citation
capability
during
the
product’s
development.
Every
answer
includes
the
source
records
behind
it,
so
that
managing
partners
can
trace
figures
back
to
the
underlying
data.
“Think
of
this
as
almost
like
a
Harvey-like
solution
for
the
business
of
law,”
Cormier
said.
“We
want
a
really
high
level
of
trust
when
they’re
looking
at
this
data.”
“AI
in
legal
software
has
to
earn
trust
before
it
earns
adoption,”
Cormier
added.
“We
are
not
asking
firm
leaders
to
trust
a
black
box.
We
are
showing
them
both
the
answer
and
the
source.”
Custom
Knowledge
Base
Another
notable
feature
is
the
ability
for
firms
to
build
an
internal
knowledge
base
within
Centerbase
IQ
that
incorporates
their
own
performance
standards
and
best
practices.
For
example,
if
a
managing
partner
asks
how
associates
are
tracking
against
their
monthly
goals,
the
system
needs
to
know
what
those
goals
are.
Firms
can
define
that
context
within
the
knowledge
base,
ensuring
that
answers
reflect
not
just
raw
data
but
the
firm’s
own
operating
benchmarks.
An
Initial
Pilot
Program
Centerbase
has
launched
Centerbase
IQ
in
a
pilot
program
with
its
customer
advisory
board,
with
roughly
half
a
dozen
firms
expected
to
be
up
and
running
in
the
initial
phase.
The
company
plans
to
create
a
waiting
list
for
other
existing
customers,
and
will
also
be
including
the
capability
in
new
customer
deals.
The
system
currently
defaults
to
Anthropic’s
Claude
as
its
underlying
AI
model,
though
the
interface
includes
a
model
selection
option.
Cormier
said
the
company
anticipates
offering
multiple
model
choices
and
potentially
allowing
firms
to
use
their
own
API
keys.
Looking
ahead,
Joyner
described
what
he
called
a
“fourth
act”
for
the
platform,
which
will
add
proactive
agentic
AI
that
surfaces
stories
and
insights
before
users
even
ask,
and
then
uses
Centerbase’s
workflow
engine
to
kick
off
actions
automatically
—
combining
agentic
AI
with
human-in-the-loop
workflows.
“It’ll
be
proactive
agentic
AI
that
will
surface
these
stories
within
your
firm
or
different
things
going
on
within
your
firm
before
you
even
uncover
them,”
Joyner
said.
The
company
also
envisions
extending
the
data
pipeline
beyond
Centerbase’s
own
data
to
include
information
from
strategic
partners
and
integrated
systems
such
as
NetDocuments
and
Billables.
Centerbase
is
demonstrating
Centerbase
IQ
at
the
ALA
conference
at
Booth
231.
For
more
information,
visit
centerbase.com/IQ.