Latest / Contractor Business Builder Podcast / AI Takeoffs Explained: The Future of Bidding for Contractors with Shiva Dhawan (Beam AI)

Transcript

AI Takeoffs Explained: The Future of Bidding for Contractors with Shiva Dhawan (Beam AI) · Contractor Business Builder Podcast

• 0:04
  Another great episode of the Contractor Business Builder
• 0:07
  podcast. Most contractors know the
• 0:08
  feeling. Plans hit the desk, timeline
• 0:12
  shrink, and your team is stuck grinding through takeoffs while
• 0:15
  bids slip away. BMI changes the diet that
• 0:18
  dynamic by automating the quantity work so your team can
• 0:23
  focus on higher value decisions and closing more jobs.
• 0:26
  So today I'm joined by Shiva Dawin, Co-founder and CEO of
• 0:30
  Attentive AI with their software BMI, which helps contractors do
• 0:35
  better take offs and then quicker.
• 0:37
  So how you doing today Shiva? I'm doing great, Derek.
• 0:40
  It's great to be here. Thanks for the invite.
• 0:43
  Perfect. Yeah, excited to get into this.
• 0:45
  I think everybody in the contracting world has been
• 0:47
  waiting for this day when AI can start to handle some of that
• 0:52
  kind of tedious takeoff and estimating work.
• 0:55
  So what was it that made you realize that something needs to
• 1:00
  change in in this side of the construction industry and and
• 1:03
  get started with creating an AI that could handle it?
• 1:08
  So the first time I came across estimators was at a conference
• 1:12
  in San Diego where a couple of estimators walked over to our
• 1:15
  booth and I was working on an AI solution that was not serving
• 1:19
  estimators. And they came to us and they
• 1:22
  said that, hey, what you guys are doing with AI and images, if
• 1:26
  you can apply that to our use case and do takeoffs for us,
• 1:32
  that'll be really helpful. So that was the first time I had
• 1:35
  lunch with them. I understood what an estimator
• 1:38
  does and I felt that this is hugely flawed because the amount
• 1:44
  of time that estimators are spending measuring plans and
• 1:49
  taking off quantities from plans in this day and age is massive.
• 1:56
  And that shouldn't happen. Like they were wasting weekends
• 2:00
  just taking off quantities. Whereas they should be focusing
• 2:04
  on increasing the increasing more time, spending more time
• 2:09
  with GCC, spending more time with suppliers, spending more
• 2:13
  times on actually value engineering the project.
• 2:18
  And like 90% of their time was going on measuring quantities,
• 2:23
  which in today's day and age where AI is so prevalent, Yeah,
• 2:26
  I can easily do that. And that's when I realized that
• 2:29
  nobody is actually focusing on this problem and solving this
• 2:34
  problem in a focused way. So that's when I conceived the
• 2:39
  idea of BMI and decided to focus on this problem because it has a
• 2:44
  huge market and it is a huge problem.
• 2:46
  And clearly in the past two years, with the growth that
• 2:50
  we've seen, we've succeeded in solving this problem for
• 2:53
  estimators in the construction industry.
• 2:56
  Yeah, that's fantastic. And yeah, so much time goes into
• 3:00
  just finding measurements and and takeoffs from plans that the
• 3:06
  information is already there. You just someone needs to
• 3:09
  measure right. And typically that's been more
• 3:11
  of a, a manual process in the past.
• 3:13
  And even with take off software, it's, it's still tedious going,
• 3:18
  you know, measurement by measurement and, and things like
• 3:21
  that. So what is the, what is the
• 3:26
  improvement that you see with going towards an AI?
• 3:31
  You know, let's say I've got a company, we've got a few
• 3:33
  estimators. How much more effective is an
• 3:36
  estimator? Or how should they be thinking
• 3:37
  about that? Like how is is this assisting or
• 3:41
  improving the the company? Sure.
• 3:46
  So there are two aspects to it. There is obviously the company
• 3:49
  aspect and then there is the human aspect.
• 3:52
  I'll talk about the company aspect first.
• 3:54
  So as a company, if you automate your takeoffs using BMI, you get
• 4:00
  the advantage of more bids out of the gate.
• 4:04
  So our customers have experienced the top of the
• 4:08
  funnel, which is the number of bits that they send out
• 4:11
  increasing by three to 4X with Beam AI coming in.
• 4:15
  Because earlier the estimators simply didn't have enough time
• 4:18
  to bid on certain projects and they were just leaving those
• 4:23
  projects, not because they didn't have the time to bid on
• 4:27
  them. So that's the first big benefit
• 4:29
  for the company. The second big benefit for the
• 4:32
  company is that we have an AI plus human in the loop solution
• 4:37
  where the human in the loop verifies the quantities of the
• 4:41
  takeoffs that we send them. So this leads to higher accuracy
• 4:45
  of the bid as well. If they do their takeoffs
• 4:50
  themselves, sometimes they are inaccurate because they simply
• 4:53
  don't have enough time and even a 2% error on a $10 million job
• 4:59
  leads to over 200 K loss, which might mean you know, their whole
• 5:04
  whole margin on the project is wiped off.
• 5:08
  So just summarizing as a company, this benefits them
• 5:11
  twofold. 1 is the number of bits that are getting out of the
• 5:15
  system they increase. And 2nd, the accuracy of the bid
• 5:19
  itself, the accuracy of the estimate.
• 5:21
  That is spot on. But the biggest benefit that we
• 5:25
  see is the human aspect of it. Estimators earlier used to spend
• 5:31
  weekends doing take off. They didn't have any work life
• 5:35
  balance. They were fed up of their jobs
• 5:37
  and as a result the in construction industry was seeing
• 5:41
  huge churn especially with the younger estimators who wanted to
• 5:45
  do something more worthwhile. And that's another thing that we
• 5:48
  are tackling. We are increasing the happiness
• 5:51
  index or the job satisfaction of estimators and that as a result
• 5:55
  is overall benefiting the company as well.
• 5:58
  Hope that makes sense. Yeah, definitely.
• 6:01
  Yeah, there's, there's a lot of benefits and, and additional
• 6:06
  things that the company can see from this.
• 6:08
  And having happy estimators is definitely a bonus for
• 6:13
  longevity. I know it is a struggle and can
• 6:16
  be very challenging to, to find good estimators who are, you
• 6:21
  know, happy doing what they do and can, can kind of crunch
• 6:24
  numbers and, and get in those nitty gritty details, you know,
• 6:28
  days on end. So by having an AI assistant to,
• 6:31
  to take care of that, it's definitely important.
• 6:33
  Now you mentioned human in the loop.
• 6:36
  Can you elaborate a little bit about that?
• 6:38
  What like how does the human and the AI trade off?
• 6:42
  Who's double checking? How do we know that things are
• 6:45
  accurate? Maybe can you explain that part
• 6:49
  of it? So initially we just built the
• 6:53
  AI and we gave the estimators an AI based solution.
• 6:58
  Now on some plans it was very accurate.
• 7:00
  On some plans it wasn't accurate, but the estimators
• 7:04
  didn't know on which plans it had it was accurate and which
• 7:07
  plans it wasn't accurate. In fact, a lot of our
• 7:10
  competitors which are pure AI softwares there, the estimators
• 7:16
  were facing the same problem. Hey, how much do I go into the
• 7:19
  estimate and see how correct that software accuracy is?
• 7:26
  And that ends up not solving the issue because again, they're
• 7:31
  spending time on looking at the quantities, looking at the
• 7:35
  takeoff, whereas we don't want them to spend any time looking
• 7:38
  at the quantities. So that's where we came up with
• 7:41
  the human in the loop solution. We said that, hey, if we have
• 7:45
  quality assurance folks on our and who can do the quality
• 7:48
  assurance and the estimator sitting in the construction
• 7:51
  company can get the final outcome.
• 7:55
  Eric, we are now living in the outcomes era where people don't
• 7:59
  want to use software, they just want software to do the entire
• 8:03
  job. And that's what B May does for
• 8:05
  them. It gives them the entire
• 8:07
  takeoff, which is 99.99% accurate and they don't have to
• 8:12
  spend a minute looking at the quantities themselves.
• 8:16
  They directly upload the quantities into the estimate and
• 8:20
  bam, the bid is prepared in a few minutes.
• 8:23
  So that's where it saves a lot of time.
• 8:26
  More than that, it saves the pain for them to go into the
• 8:29
  takeoff and looking at the quantities of each and every
• 8:32
  takeoff. Yeah, very important point.
• 8:35
  I think that's kind of one of the fears or uncertainties of
• 8:40
  contractors using services like this even before, you know, I
• 8:46
  could do it. But even just having an external
• 8:49
  estimator, there's always that thing in the back of the mind
• 8:52
  of, you know, did they put 1000 square feet instead of 10,000
• 8:56
  or, or or something like that, right?
• 8:57
  Or even just small amounts out. Margins are so thin, especially
• 9:01
  on large commercial projects, that off by a couple percentages
• 9:05
  can mean a big difference in in the, you know, market profit
• 9:07
  margin at the end of the at the end of the job.
• 9:11
  So, yeah, good to hear that you've directly applied, you
• 9:18
  know, human in the loop to, to ensure accuracy and and as AI as
• 9:24
  it continues to get better. So just as you know, there may
• 9:28
  be some accuracy limitations especially for newer type plans
• 9:31
  or more complex stuff or whatever, that training will
• 9:35
  continue to get better right into the future and and get more
• 9:38
  and more accurate. Now it's exactly types of
• 9:40
  contact. Go ahead.
• 9:43
  I just wanted to add there that we are already seeing that the
• 9:46
  amount of work that the humans are doing on our end is reducing
• 9:50
  drastically over time because the air is becoming much more
• 9:53
  accurate. But from a customer lens, from a
• 9:56
  contractor lens, the fact that they know that a human has
• 10:00
  eyeballed it, they feel much comfortable and they are very
• 10:06
  sort of satisfied with the results, even if it takes a
• 10:09
  little more time for them. Yeah, definitely.
• 10:14
  Now what type of contractors do you feel get the best benefits?
• 10:23
  Are you like have the easiest time using the AI?
• 10:26
  Is there certain sizes of company or locations or like
• 10:31
  trade contractors versus general contractors?
• 10:35
  Who do you see having the most success with with us?
• 10:40
  Derek, when I started the company and the product team, I
• 10:44
  I knew that I wanted to serve all the contractors and people
• 10:47
  came and told me, hey, don't boil the ocean or don't go after
• 10:50
  anyone. Everyone, you lose focus, but I
• 10:54
  was like, hey, everyone wants take offs, right?
• 10:56
  So everyone should be able to get the solution.
• 11:00
  So right from general contractors, which contributes
• 11:03
  to what 20% of our revenue, 2 subcontractors, which includes
• 11:09
  MEP contractors, concrete and rebar contractors, civil
• 11:12
  contractors, utility contractors.
• 11:15
  We have customers across the board and for some customers
• 11:20
  like AMAP contractors, we are now also offering a
• 11:24
  do-it-yourself service which is which does not have a human in
• 11:27
  the loop. Because some contractors came to
• 11:29
  us and they said, hey, I don't want to wait for two days with
• 11:33
  the human in the loop output. If you give me the raw API
• 11:36
  output, it's good enough. So for them, for MEP contract
• 11:40
  serving demo hybrid solution, which has the human in the loop
• 11:45
  solution, but also the raw AI outputs in case they need it.
• 11:49
  We're just stepping back across the board from general
• 11:54
  contractors to Subs. We serve all of them because all
• 11:57
  of them need takeoffs. OK, Yeah, makes sense.
• 12:03
  And let's say I I've got a contracting company, we've got a
• 12:11
  maybe one to three estimators. What is, what is the
• 12:15
  implementation look like getting started?
• 12:17
  What's, what's the process there and what are the expectations
• 12:20
  over the first, you know, month or two?
• 12:24
  Absolutely. So typically before we sign the
• 12:29
  contract, we do 2 pilots as in two plans where we process the
• 12:36
  pilot and we show the contractor that hey, this is how the raw
• 12:41
  output will like the end output will look like.
• 12:45
  That typically takes two to three weeks.
• 12:48
  After two to three weeks, we signed the contract and then
• 12:52
  comes the onboarding phase where we tell them, OK, now we need to
• 12:56
  do, we have a rough idea of what you need, but we still need to
• 13:01
  do more takeoffs for you, at least five more takeoffs.
• 13:04
  And we also want to integrate with whatever estimating
• 13:08
  software you have. So this typically takes another
• 13:12
  three weeks. So right from the point of the
• 13:17
  first touch point where somebody books a demo on the BMI website
• 13:21
  or calls us to book a demo from there you can take roughly one
• 13:26
  month and BMI is live in the in the in the company.
• 13:32
  So it's not a huge implementation cycle at all.
• 13:37
  It's fairly plug and play. And that's what I tell all
• 13:41
  contractors that, hey, if you try a new AI software, you might
• 13:46
  have that initial inertia of you know, how it will work, how it
• 13:50
  will help my team. But if you just take it for a
• 13:53
  test drive, it won't take a lot of time and you'll quickly get
• 13:57
  used to it and overcome that initial apprehension or inertia
• 14:02
  that you might be thinking of before you start it.
• 14:07
  Hmm, that's great. I like how you have customized
• 14:11
  approach as far as doing specific takeoffs for a company
• 14:15
  and, and learning, you know their needs and, and what types
• 14:18
  of projects, plans, etcetera that they that they work on and
• 14:21
  and then helping them implement throughout their team.
• 14:27
  Is there any maybe mistakes or or missed thoughts around
• 14:34
  utilizing a software like this? If companies like a little bit
• 14:37
  apprehensive or maybe the owners on board, but maybe the
• 14:40
  estimators aren't too thrilled about it.
• 14:42
  Have you have you seen anything like that or any mistakes or or
• 14:46
  pre planning as far as like on a leadership state stage to to
• 14:50
  tell like to bring this into a company?
• 14:54
  Yeah. That's, that's a great question.
• 14:58
  I feel like some of the initial mistakes that we made were that
• 15:05
  the owner or say VP in the company really wanted the SL
• 15:10
  Musion, but the estimator was not on board.
• 15:14
  So they signed the contract. You know, they did everything
• 15:17
  but the user who's the estimator, When they actually
• 15:22
  started using it, they were not comfortable with it.
• 15:26
  In fact, some of them feared that, you know, they'll lose
• 15:29
  their jobs. So that's I think the biggest
• 15:33
  misnomer that I sort of clarify. 2 estimators and I think all of
• 15:39
  us should like think about this because with AI, a lot of people
• 15:46
  think that, you know, they might lose their jobs.
• 15:49
  But that's actually not true because if you look at history
• 15:53
  of the last 100 years, right, there have been so many
• 15:56
  innovations like when the ATMs came up like bank tellers
• 16:02
  thought that they will lose their jobs.
• 16:03
  But actually, the number of jobs in the banks increase because
• 16:08
  like there was more work that needed to be done, right?
• 16:13
  So when these digital innovations come, and it's true
• 16:17
  for AI as well, it's true for Beam AI as well, Everyone and
• 16:22
  including the estimators should think of it as, hey, now this is
• 16:27
  making me more productive. It's actually changing the way I
• 16:31
  used to do work. And that's where some of the
• 16:34
  insecurities might come in. But if you look at it from a
• 16:38
  positive lens, that's actually making them do more work, do
• 16:42
  more bids. And I, I don't think in the last
• 16:47
  two years we have 1200 contractors and I don't think in
• 16:52
  the last two years any of the contractors have fired their
• 16:56
  estimators. In fact, most of them have
• 17:01
  upskill themselves from an estimator who was a pure
• 17:05
  quantity surveyor to now an estimator who's like a precon
• 17:09
  specialist and doing multiple parts of the precon value chain
• 17:13
  instead of just calculating quantities.
• 17:17
  Yeah, that's fantastic. And good thing to articulate and
• 17:22
  clarify because that is another kind of fear of AI in general.
• 17:26
  But in this case it would be very applicable to estimators is
• 17:30
  this isn't a full replacement. This is like an assistant and I
• 17:34
  think most employees in a company would love to have their
• 17:38
  own assistant that they could hand off work to and not have to
• 17:42
  do. Some of them are like, you know,
• 17:43
  tedious or repetitive type stuff and and AI is great at that kind
• 17:47
  of stuff. And and then, as you mentioned,
• 17:50
  their role can actually expand where instead of having as much
• 17:53
  time crunching numbers, they can now focus more on
• 17:56
  preconstruction or supporting the rest of the team or, you
• 17:59
  know, fulfilling other pockets in the company where where, you
• 18:03
  know, good human insight is needed.
• 18:04
  So, so that's fantastic. Now for what contractors out
• 18:08
  there listening and watching, what's the best way to find out
• 18:12
  more about BMI and and get a demo?
• 18:16
  So for all contractors, you can just log on to our website,
• 18:20
  www.ibeam.ai and book a demo from there.
• 18:25
  And one of our reps will reach out to you and they'll set up a
• 18:29
  demo with you and go take you through the entire process.
• 18:35
  And yeah, that's as simple as that.
• 18:38
  Wonderful. So we'll put links in the
• 18:39
  description. Go click the link, go check out
• 18:42
  Beam, get get hooked up with their team there and stop
• 18:45
  estimating the old, old boring way and get some AI of working
• 18:49
  for you and your business. Thanks so much for coming on
• 18:51
  today, Shiva. Thanks, Derek, It was a
• 18:54
  pleasure.