00:10thank you so much for having me here you
00:13know sunny here gives me so many
00:15memories you know just walking through
00:17Huang remembering my teaching assistant
00:20classes and and I brought back my
00:23memories of when I was a grad student
00:25here at Stanford now Stanford's not just
00:28special to me because I'm not I'm an
00:31alumni here but also because we started
00:33one concern right here four years ago it
00:37was here where I met Tim and aha with my
00:40co-founders where we eventually became
00:43friends and he became friends because we
00:45shared common values if you gone through
00:48our demographics for the co-founders we
00:50are very different my age I'm an Asian
00:53hermits Muslim Indian and Tim's
00:56Caucasian from the military but you know
00:59we had fun at Stanford always talking
01:01about large problems and just really
01:03getting to know each other
01:05now one fine summer of 2014
01:08Mohammed went back home to Kashmir and
01:10if I'm not sure if he know about this
01:13but Kashmir India had one of the largest
01:15floods that here a flood which they
01:18hadn't seen in several decades we tried
01:21very frantically to reach out to him we
01:24would see the news talking about how
01:26many hundreds of people ended up losing
01:28their lives and we couldn't reach out we
01:31couldn't connect with him and was very
01:33very scary ultimately we did end up
01:36connecting with him to figure out that
01:38his family and him were safe after the
01:40flood however when he came back to
01:44Stanford he talked about so many stories
01:46related to the Kashmir flood and it was
01:49pretty modifying it was a disaster in
01:52all sense K there was complete chaos
01:55after the disaster you would have no
01:58sense for the emergency rescue to figure
02:01out where they should actually go who
02:03should they actually rescue he had
02:05thousands of people abandoned on the
02:08rooftops with me for several days and
02:10you have hundreds of people either
02:12drowning or being washed away by the
02:15flood and we wondered why is this still
02:18a problem and and wise in the 21st
02:20century why is nobody thinking about it
02:23what can we do to help just being
02:26graduate students here a little while
02:30later because the opportunity to work in
02:32a class project together it was the
02:35machine learning project at Stanford as
02:37well as the probabilistic earthquake
02:39engineering class at Stanford and we we
02:42took that opportunity to figure out can
02:44we actually solve that problem we saw in
02:46the summer of 2014 and and what can we
02:50do to reduce that chaos the goal there
02:53was just you know let's find a way to
02:55figure out what exactly the problem was
02:57and the problem we try to resolve was
02:59can we provide a real-time granular
03:03understanding of which buildings and
03:05which people are in what state of
03:07collapse during a disaster
03:08so that was a problem statement our
03:11backgrounds of computer science
03:13structural engineering and earthquake
03:14engineering led to it being focused
03:17mostly on seismic initially and and we
03:20were very happy with the results it was
03:23surprising cause it wasn't done before
03:26but there's a lot of physics and science
03:29already there in structural and
03:31earthquake engineering all we did was
03:32sort of understand the features present
03:34there and really convert that in a more
03:36granular and real-time solution we're
03:41happy with the algorithm we created but
03:43what we were more pleasantly surprised
03:46was that when we presented this
03:48algorithm in the in the in the CS 229
03:52machine learning fair a lot of
03:54professors came up to us a lot of
03:56investors came up to us and said that
03:58what you have is applicable and you can
04:01create a real product here now we were
04:05taking a little aback because we never
04:07thought of starting a company or of
04:10creating a product and none of us had
04:13experience running a business before
04:16two-thirds of us were immigrants myself
04:19an argument and I'm not sure whether
04:22it's similar thought process but when
04:24you're immigrant generally you come to
04:26Stanford you know your thought process
04:28is work hard study hard get good grades
04:31and then fines are a putable place to go
04:34to so that you can then give
04:36back to your family who sacrificed so
04:39much to just make sure that you come
04:40here so so when somebody came up to us
04:43and said you know you should not think
04:45about the idea this way think about what
04:48our product could look like it was a
04:52little hard for us because that meant
04:54that there is obviously going to be risk
04:56associated with creating a company and
05:00we didn't have any background associated
05:03with that so what we insert there is we
05:07try to see whether what we're hearing
05:10from our professors and from from the
05:12investors is that really real so so we
05:15cold called a bunch of different cities
05:17a bunch of different emergency officials
05:21and we ask them that hey you know we are
05:24three clad graduate students here we
05:26have this algorithm we really don't know
05:28whether or not it would be helpful or
05:30not can you spend a few hours with you
05:32to just understand is this really a
05:34problem and can we can we do some impact
05:37here and we're surprised that instead of
05:40them spending a few hours they ended up
05:43spending several days with us the hope
05:47and the excitement we saw in these city
05:49officials I think that was the final
05:52push to make us say that you know we
05:54have to do this it sort of came to light
05:58that it is a responsibility we and if we
06:01don't do it will anybody else think
06:04about this problem or will this problem
06:05keep repeating itself so then all the
06:09three of us the three co-founders we had
06:11a really hard conversation at least
06:13actually several hard conversations
06:15about what we would have to give up in
06:17terms of our previous ideas and what
06:20does it mean to commit to to a mission
06:23and what does it mean to commit to a
06:24company now while I talked a little bit
06:27about you know my aunt Ahamed as
06:30immigrants you know are our problems
06:32tim himself was raising her family so
06:35you know what would that mean in terms
06:37of his responsibilities but at the end
06:42of those conversations we said you know
06:45we are all in you know we have to do
06:47this and once once we figure out that
06:50is the problem you want to solve we then
06:52started thinking about the next step the
06:55next step is you know parallely
06:56continuing our conversations on what the
06:59product would look like with the cities
07:00but also figuring out that we cannot
07:04build a very complex machine learning
07:06pipeline for disasters with just the
07:09three of us so building a team building
07:12something actually feasible so that
07:15meant trying to figure out do we want to
07:17actually do this on our own raise money
07:20ourselves or start looking at investors
07:23it was hard for us to pitch in money
07:25ourselves you know being immigrants and
07:27all and so that meant that we did have
07:28to go through the investor out now now
07:33that was a little surprising as well
07:35because when we went to investors we
07:37thought you know obviously these cities
07:39are so excited you know we heard about
07:41product market fit you know customers
07:44didn't like it so you know investors
07:46should definitely jump on this idea and
07:48you helped us through but that wasn't
07:51what we saw for around nine months we
07:55mostly heard rejections and it's very
07:58hard to hear that because along with the
08:01risk you're taking you hear somebody
08:03really experienced with a lot of
08:05expertise and building a business coming
08:08to you and say saying no you have a
08:11noble idea but it's just an idea and you
08:14need to give it up I have an anecdote of
08:17a particular investor which we were
08:19really you know we really respected and
08:22that was one of the meetings which I
08:23still remember we went into the meeting
08:26we gave our pitch tag and and the
08:31investor said you know if an earthquake
08:33happens I'm just gonna walk out and go
08:36to my neighbor's house this is such a
08:38silly idea and you should start thinking
08:40about something else completely and you
08:42wasted so much time so nine months of
08:45not or just hearing rejections nine
08:48months of not knowing basically
08:51collecting debt and then nine months of
08:53real big fear from your visa process you
08:56know what's going to happen can i still
08:58stay here do i have to go back to india
09:00it was pretty scary for
09:03but instead of having that bring us down
09:06what we did was we took a step back we
09:10said you know what are we not doing
09:11right we did we didn't blame the
09:13investors or the feedback we realized
09:16that there's something missing in what
09:18we are doing which is why we couldn't
09:20portray our story to to the investors
09:24while talking to us and so we later
09:28realize that the problem was twofold one
09:32problem was we weren't talking to the
09:34right set of investors you know they are
09:37different investors with different
09:38investment thesis so they're few in the
09:41seed stage who definitely want to look
09:43at evidence in the market you know put a
09:45strong ratio or weight on that as part
09:49of investment in your company
09:50and given that there was no earthquake
09:53gov tech product out there you know just
09:56really high risk whereas there are other
09:59investors who put a lot of weight on the
10:03team is there need you know and that the
10:06team's gonna figure it out like really
10:08large focus there so we realized that we
10:10need to start looking at you know these
10:12investors the second thing we realized
10:15was what was missing in our pitch was we
10:18always thought about the vision but very
10:20little about the feasibility of the
10:22business and we knowing that we didn't
10:25know that too much we took whatever help
10:28we could get we went to Stanford
10:31adventure studio are talking to all the
10:33Stanford graduate students they're
10:35pitching several hundreds of times and
10:37and just brainstorming about what could
10:39be wrong we talked to investors for
10:42advice even the people who rejected us
10:44telling them of what could we have done
10:46better what is actually the problem and
10:48we we invited all that information in
10:51and an hour ago eventually finding the
10:55right investor and helping them
10:57understand exactly the perspective of
10:59why this idea should be present was what
11:03founded one concern so fast forward
11:07today we crew from the team of three to
11:11a team of eighty and our mission is for
11:16where we see resilience in the three
11:19pillars of safety equity and
11:22sustainability and we obviously expanded
11:27from seismic to floods and fires and to
11:30talk about what the issue in seismic is
11:32when an earthquake happens right now
11:35what what what is in the hands of
11:38emergency responders there's two things
11:41you might get one is a map where in
11:43entire neighborhoods or cities a color
11:46are colored a single color you know red
11:49yellow green telling them that that is
11:51the impact for a city the second thing
11:54is you probably seen it on Google it's a
11:57shake map you see really large red
11:59concentric circles on the map and that's
12:02sort of what a responder is supposed to
12:05take in and and do those life-saving
12:07decisions so when you think about why
12:09the chaos follows after it's obvious you
12:12know just looking at a blurry map how
12:14can a first responder know what to do
12:17and when when several thousands of
12:18people need help so we moved from that
12:21idea into the idea on the left really
12:25focusing on going really granular on a
12:28block by block level and even building
12:30level helping people understand which
12:33which buildings are on what sort of
12:35collapse minutes after the earthquake
12:36happens we not just do building level
12:40information we understand understood
12:43that the people component is very
12:44important so who resides in those
12:46buildings are they low-income are the
12:49senior population are the children
12:52because the resources you spent you send
12:54during a disaster are very different
12:57based on the vulnerable population who
12:58are affected Hurricane Katrina is an
13:01example where in most of the there were
13:04a lot of senior citizens who had
13:05abandoned and they needed an influx of
13:08blankets because they were shivering
13:10during the hurricane but no one knew
13:12exactly what was the ratio of people who
13:15are affected so that's the second
13:16component we focus on in in a product
13:20and finally the third component is we
13:24don't want to just predict which
13:26buildings are down all we want to help
13:29is which what is the state of the city
13:31so that implies not just looking at
13:34direct causes of impact but also looking
13:37at secondary and tertiary causes of
13:40impact so is is a power sector down is
13:44your healthcare down is your water
13:45system down and how does that ultimately
13:48go and affect citizens give an example
13:51you might have the strongest Hospital in
13:54the world which might be up during the
13:57big one and during a big earthquake but
13:59if the Hetch Hetchy system is down or if
14:02the power system is down and there's no
14:04sort of back-up plan it still means that
14:06your health care is affected it still
14:08means that citizens aren't getting that
14:10rescue so how do you protray a cascading
14:13effect of dependencies along with the
14:15impact map to the first responders flood
14:21we just released last November and we
14:24chose floods specifically because it has
14:27a larger impact and it is definitely the
14:31impetus of one concern the Kashmir flood
14:34it's more or less similar to what we do
14:36in seismic the only difference is that
14:38you could actually forecast five days
14:42now when hurricane Harvey happen or any
14:47hurricane happens what data these cities
14:50understand is a storm is coming it's
14:53going to come in three days
14:54it gonna get five inches of rainfall but
14:56how does that help you know who to
14:58evacuate and how does hazard help you
15:00know who is going to end up drowning in
15:03that particular incident so it's a pride
15:05providing that really granular impact
15:07and of good and updating your accuracy
15:11as and when you get real-time input from
15:13on the ground we are very transparent
15:16about how accurate or inaccurate our
15:18models are in fact that's what cities
15:21love they want to know how can they help
15:23us make our models even more accurate we
15:26want them to be able to make those
15:28decisions after understanding where our
15:31models wouldn't work and where our
15:32models will work in the final third
15:35component which is still in development
15:37is fires fires are very close to
15:40California the costliest is
15:43sort of last year was the wildfire in
15:45California now the issue of impact
15:50knowledge and situational awareness is
15:52even la is probably larger or or equal
15:56to that of floods and fires floods and
15:59earthquakes you have fires moving at 80
16:03miles per hour literally engulfing
16:05football fields in seconds and meanwhile
16:08the only data which is coming in is 911
16:10calls or the recognitions where you're
16:12driving around the neighborhood how does
16:15that help you really understand what the
16:17state of impact is and how does that
16:20have first responders or firefighters
16:22really evacuate ahead of time and and so
16:27that's sort of our different offerings
16:30and we moved to not just doing response
16:33we understood that it's not just about
16:36during the response helping save a few
16:38lives it's definitely about how do we
16:41prevent the loss in the first place we
16:44work with cities like San Francisco Los
16:46Angeles and mult and get also in the
16:49midst of deployment internationally
16:51we're in we're in what we do is have
16:54them do more efficient planning how do
16:56you prevent this loss in the first place
16:58one is just through looking at multiple
17:01different scenarios seeing is your
17:03emergency response system actually
17:06capable of responding to it
17:08do I need to increase my budgets do I
17:10need to add more resources do I need to
17:12communicate better so that's what we
17:15call a scenario planning but the second
17:18thing is about reducing the loss in the
17:22so that's through by the city planning
17:24that's through better infrastructure
17:26adjustments just trying to figure out
17:28even addition of you know that Hetch
17:30Hetchy water pipe which would have you
17:33know cause a lot of chaos to the high
17:35health care in California so how do you
17:37like really prioritize among all those
17:38decisions and we're doing this not just
17:43for governments we understanding and
17:46helping people work together in our
17:49resilience ecosystem wherein we get in
17:52different commercial sectors and
17:53governments talking the exact same
17:55picture during a large event most of the
17:58small medium businesses will absolutely
18:00go bankrupt so how do you have people
18:03understand that this is the state of
18:04what our disaster risk is and how do I
18:07prepare for it ahead of time so what
18:11what we do here is you know we call this
18:14benevolent intelligence where and we
18:16really want to use artificial
18:18intelligence for social good and we
18:22really think that we can enable a
18:24disaster-free future and what I mean by
18:27this Astor free I mean that hazards will
18:30keep happening you'll have seismic
18:31events you have floods you'll have fires
18:33but cities and citizens will be able to
18:36immediately bounce back and that
18:38bouncing back would happen because of
18:40immediate response during a disaster
18:42through better planning before a
18:44disaster and even through you know
18:47better infrastructure updates and it's
18:52it wasn't so that's that's one concern
18:55and they were we we come here like I
19:00said we've grown pretty large but it
19:02wasn't very easy coming to where we are
19:04they they've definitely a lot of
19:07challenges we had to face and so the two
19:12different challenges I'd like to focus
19:14on is number one is surround yourself
19:17with the right people
19:18it's always going to be about the people
19:21you have the right people
19:23I think 75% is already half done and
19:26what do I mean by people I mean by
19:30obviously the the team the people you
19:31hire you know we've been lucky
19:33we've hired not just divorce thought
19:36processes but like divorce in terms of
19:39age gender race etc but also in terms of
19:42thought trusses we have firefighters we
19:45have emergency managers we have mayor's
19:47we have data scientists we have a
19:50collection of a lot of different thought
19:52processes which enable us to understand
19:54how can we fix this through policy and
19:56technology and make sure definitely that
20:02the bar you'd keep for your mission
20:05alignment is as high or maybe even
20:08technical bar or the business body
20:10looking for don't compromise on that
20:12we've learned that the hard way that we
20:14should never do that I give you a quick
20:19anecdote of an interview we did in the
20:21early stages of one concern we had this
20:24really really smart engineer come to
20:27interview with us I'm really impressed
20:31so I interviewed him and one of my other
20:35engineers interviewed him and he was
20:38very rude and condescending throughout
20:40the interview didn't let me complete my
20:42sentences and I I was taken aback and
20:47after we did a debrief the the CEO came
20:51in saw Hamed and he was like no that guy
20:53was wonderful he was very nice with me
20:55he was very respectful so I don't know
20:57what was the difference after looking
21:00through a debrief notes it was pretty
21:02obvious that the person who came in had
21:04a bias towards being respectful to you
21:07know men in particular and you know that
21:11was a complete whether or not he was the
21:13most smartest engineer in the world
21:15I couldn't have somebody come in and be
21:18disrespectful to my team and this was
21:20not just something I thought of AHA met
21:23the entire team agreed that this was
21:25something you would always stand by so
21:27culture shouldn't be something light you
21:29should be willing to not hire if
21:30somebody is not your culture and really
21:32stand stand by it the second people
21:37component especially for a
21:39mission-driven company is definitely
21:42surrounding investors so during the seep
21:46town it was a little bit hard but
21:49ultimately we ended up in a series a we
21:52got a lot of interest for multiple
21:54different investors now say we we caught
21:58somebody on the board who told us that
22:00you know this is a way you could get
22:02more dollars but you'd have to give up
22:04your mission of saving lives in any
22:06sense you know that's a complete no-go
22:10for us as well and so we wanted to make
22:12sure that there is complete mission
22:15alignment with our investors and we took
22:17a lot of time with the potential
22:19investors asking them where do you see
22:22what would you do if if you are not
22:24seeing all attraction in this particular
22:26markets really just understanding how
22:28they would react to this and are they
22:30really aligned to our larger mission of
22:32saving lives and so we were pretty lucky
22:35getting the right board members on board
22:36and I do think that is something you
22:39should always insist on getting to know
22:42your board members and understanding
22:44that we are all all in sync on what our
22:49the third one which is obvious is of is
22:52is your founders the cofounders you work
22:56with so I was lucky I mean if I was
22:59friends with my previous co-founders we
23:01sing Donna values but you know you need
23:05to be I've been so vulnerable with my
23:07co-founder several times I've broken
23:10down completely I've told them all the
23:12different flaws have had and they've
23:14done the exact same thing for me when
23:16things were rough I could call either a
23:18her mother Tim at 12 o'clock in the
23:20night and freak out and let them know
23:21you know this is the problem and they
23:23would not take hold it against me they
23:25would let me take a step back and they
23:28would give me their perspective of what
23:30I could do pattern and I would always do
23:31the same for them which comes to my
23:34second challenge the emotional burden of
23:37you know running a company being a
23:41founder is going to be very lonely it's
23:43going to be very hard because you always
23:45want to be the pillar of strength for
23:48your employees for your founders for
23:51your clients for even sometimes for
23:53everybody you know you always want to be
23:55the pillar of strength who people to
23:56people go to but there's going to be a
23:58lot of ups and downs many downs more
24:01than ups generally in the start and if
24:04you are not careful it can take a really
24:06large toll on you you need to be able to
24:10take a step back and and do what I did
24:12similarly in the rejection piece which
24:14was don't take these rejections
24:16too hard critically on yourself take
24:19this as opportunities to really learn
24:21how can you take that negativity or two
24:25and convert it into something you can
24:26learn from and it's very important to
24:30find the right support system as well I
24:35my family's completely bought in
24:36initially they want and then my friends
24:39are completely bought in you know they
24:41know that I might not be able to spend
24:43six months with them because I'm working
24:46so hard on one concern but they all
24:48support me through and through
24:49so it's very important to find the right
24:51support system otherwise it will be very
24:53very hard on you if this and despite I
24:59guess but a one last component I want to
25:03talk about is don't look too much into
25:04the details when we started a company
25:07the technology was very very scrappy we
25:10just build a bare-bones web application
25:12you know actually a machine learning
25:14algorithms were on MATLAB or MATLAB
25:16server and we didn't really care about
25:18let's find the most latest technology
25:20and let's make it the coolest app ever
25:22what we'd instead focus on was let's
25:25work with the cities and really figure
25:27out what exactly should this be and now
25:30obviously that's not the case we have a
25:33very very secure a lot of micro services
25:35and we did invest in technology but
25:38don't overthink that focus on what
25:40really matters which is the problem even
25:42something you want to work on and you
25:47know despite all the different
25:48challenges I talked about I would always
25:54the excitement I see with the team I've
25:56surrounded with the opportunity I have
25:59to just work with phenomenal people to
26:01make a difference you know always will
26:03make makes it worthwhile and and you
26:08know if I didn't do this I I don't even
26:10know would this problems will be solved
26:13would anybody think about it I'm hopeful
26:16that at some point you know five years
26:18from now we're able to create at least a
26:20single metric on on a wall in the
26:22company which talks about the number of
26:24lives we affected in a positive manner
26:27and that would be completely worth it
26:29and so if there is a particular idea you
26:33have or and you really believe that it
26:35has to exist you should go for it do
26:38some research obviously the research we
26:40did with the different cities but but
26:43take that step and don't and and it as
26:46long as you are completely committed and
26:48the right team with you you know it
26:51should solve itself out so I think
26:54that's pretty much it
![Welp, Here's what the Fed *Just* Said [Fed FOMC Minutes]](/_next/image?url=https%3A%2F%2Fi.ytimg.com%2Fvi%2Fl8u23oq1E9E%2Fhqdefault.jpg&w=3840&q=75)
