learning about programming through my daughter’s eyes


i am a valley kid, born and raised. now, as a parent living in the valley, i wonder about my five-year old daughter and how she will be able to pursue an interest in computers, mathematics, programming, and hacking — if she so desires, of course. what my daughter doesn’t know is that, not too long ago, her dad (me) was hacking on computers not too far from where we live today as a kid. during that time, working on computers was far less trendy and romantic. i first taught myself basic on an apple 2+ then later on pascal. i was then taught turtle logo as a part of a special program at school, which while it felt a bit of a regression from basic i could see how it was a decent environment to teach others to program. for me, during that time, the motivation was to write games.


now, fast-forward to today, i see my daughter use an ipad effortlessly. and, i want to make sure she has access to resources and tools to experiment with, should she be so inclined to go beyond using the apps. i’ve checked out hopscotch. there is scratch from mit. there are local groups who are providing a space for kids to explore. whatever my daughter chooses, my hope is that there’s creativity — in art, or athletics, or music….or computers.


how will kids today learn how to hack? while games were my motivation (as well as breaking the security on the games in order to copy them and exchange with friends), it provided a discovery platform. there were no limits. no directions (at least from my parents). yet my friends and i would spend hours removing the copy protection and then adding a  “cracked by…” as the game loaded.


legos today have directions – legos when i was a kid did not. in our current environment, how do we maintain that environment of endless possibilities? in the near future, i believe everyone will need to be able to program – much like learning another language. obviously, there will variations of the depth of this learning – but the ability to apply logic/reason and assemble a set of commands/functions to perform an activity is great on many levels.


my daughter was also an inspiration for me to initiate the women in engineering effort while i was at twitter – which has been amazing to see grow and flourish with each passing month. so awesome! we need to get more women in computer science and math  - so i am at least going to expose my daughter to what can be done – if she so choses that path. ironically for  me – because i spent so much of my childhood on a computer i rebelled in high school a bit and wanted to check out biochemsitry/molecular biology. but over time….returned to my roots and frankly i was better at programming than primary research. the current app that i am working with olivia on an ipad with is hopscotch - http://www.gethopscotch.com - she loves the idea of building her own game. and, as she gets older, i’ve heard of slightly older kids messing around with raspberry pi projects. and, there’s this whole subculture of kids playing minecraft and hacking those games to bend to their tastes.


i’m curious, if you have young kids, how do you seem them interacting with computers, and what sites, programs, resources, or groups have you seen? what are folks doing with raspberry pi and minecraft? i’m eager to connect with folks who are interested in this and to learn more, and of course, to help come up with more and more creative ways for the youth in the area to be exposed to this new language.

the long slow flip


“whatever good things we build end up building us.” -jim rohn

in the tech world today, value is being created so fast, it can feel like it’s just overnight. it’s exciting. from palm to twitter to sand hill road, it’s an electric time to be a builder and investor, no doubt. while the pace of value creation ticks up, i do wonder what we may lose as a result, the risk that used to be associated with joining a startup — the stigma, even. i wonder if startup jobs are the new corporate jobs, as former colleague @al3x provocatively articulated. but mostly, i wonder if the long years of pain and suffering even the successful companies of old experienced may be lost on a generation.

i’m uncharacteristically nostalgic today because i have some uncharacteristic news to share: a company that i founded in 2002 — 11.5 years ago — is finally going to have an exit (and an exit that provides a strong return for investors and a great outcome for colleagues and team members). back in 2002, i had a simple premise that the standardization of interfaces to data would enable a new layer of processing to be created — data virtualization. thankfully, i convinced other people the idea had merit: the engineers, investors, executives, and customers that would come to form the nucleus around the company. today, i’m excited to announce that composite is partnering with cisco to take data virtualization to a grander scale. building composite was the roller-coaster everyone talks about, but it was also a blast, mostly because of the people involved. since 2002, composite “alums” have gone on to big things, such as founding or being on the starting team of Adchemy, Marin Software, Ludic Labs (Groupon), Reverb, and Hortonworks, to name a few. saying “thank you” on this blog feels a bit inadequate given the blood and sweat of the engineers who worked out of my house to build the initial version, to the many advisors and friends who helped me as a first-time founder/CEO, to the investors who believed that in 2002 startups could still deliver venture-scale returns, to jim green who left his public post to become ceo when i wanted to return to doing technical work, to stephen ahuero/will lam/wilson yu who have worked at composite since it started in my house and are still employees today, to the countless employees who helped build a real enterprise business that generates real, sustainable value, and to my friends….i will be forever grateful.

i normally wouldn’t boast about an exit, as it’s not my style, but the main purpose in writing this is for the founders and future founders out there — to point out that i started composite software almost twelve years ago — 12 years! — and finally today got to the elusive exit all founders seek. it didn’t happen in one year, or within a venture cycle, or even within a decade :-)….it took a lot of sweat, patience, low moments, fear, drive. there’s a great post by ben sesser on pando, “it’s damn hard to build an enterprise business.” go read it. it strikes a deep chord with me. and it takes the commitment of scores of believers — not just the founder — to carry the company through. while it all may sound cliche, i’ve lived it, and it’s true. we all have the scars to prove it, which is why today tastes so good.

my friend and mentor: rajeev motwani


four years ago today,  rajeev motwani passed away, but his spirit lives on.
a few weeks ago, i was emailing with a friend in the valley, about mentorship and our times through the valley. we may be young, in one sense, but we’ve been around the valley and technology now for some time. as we discussed mentorship, my friend asked a curious question: “who filled the hole once filled by rajeev?”
i didn’t have an answer. there isn’t an answer. there is only one rajeev. 
the “rajeev” he meant was rajeev motwani, famed professor, mentor, investor, and advisor to some of the valley’s biggest and brightest computer science minds. like many others, i was lucky to not only call rajeev a friend and mentor, but had the added luxury of having him as an investor and advisor in my first company, composite software. it was funny how we met — when i was raising money for composite, three different venture firms asked rajeev to conduct tech due diligence on my work.
i was intimidated of rajeev’s intellect at first, but as we spent time together, rajeev invested in me and, in the process, i got to know him and his family very well. he was a very trusted advisor. he helped me think through financings. he encouraged me not to bring in an external ceo. he also encouraged me later on to move on from the company, as i’d lost control of composite and needed to move on. those were tough decisions, but having rajeev there made them much easier. i went up to microsoft, and rajeev connected me with great folks there. when i’d come back to visit, i’d always see him.
a good number of people reading this will know rajeev and probably nod their head as they’re reading this. they’ve felt his mentorship and, in some way, still do. and, there’s a younger set of you who will have to google him and read his wikipedia entry. but, those entries don’t do him justice. he was taken from us too young, and i still think about him when i’m driving home on el camino. and, i wonder who does, in fact, fill rajeev’s shoes these days? who is out there mentoring for the sake of giving back, for the sake of pushing talented people to the limit, to achieve their potential and be there when things aren’t rosy as it all seems. can anyone fill his shoes, let alone just one toe?
how did rajeev have all this time to help me and countless others? i’m trying to fill in a small part of the hole left by rajeev, just a little. it’s impossible to do so, to the level he did. i’m hoping more of my peers do as well. and i’m hoping that, somewhere, out there in the world, the next rajeev — if that’s even possible — graces us in the valley with his or her presence. these days, we need it more than ever.

observations from office hours@stanford

i am going to try to increase the frequency of posts to my blog for the rest of the year (2013 has had a slow start!) so feedback on topics etc is welcome!


for the past year, i’ve had the good fortune of conducting office hours in the cs dept at the gates building on the farm. once a month for an afternoon, i’ve been hanging out with swarms of students. while it may appear that they are asking me for knowledge and information, the truth is that it is me who is learning from them — and, learning a lot!
like many of you, over the holidays a few months ago, i read the brilliant medium post by branch co-founder josh miller, in which he catalogued some of the observation his sibling (who is a teenager) made regarding consumer tech products. in that same spirit, inspired by josh’s post, i wanted to share the trends i’ve observed in spending lots of time with engineering students on campus:


  • video is in its early days: for these college kids, firing up a video on any mobile device happens all the time, even to listen to music via youtube. i do this on my laptop when i’m working and/or want to listen to a specific song or catch up on on a technology video talk, but the manner in which they discover, share, and communicate about videos they find — from every content vertical — is something many of us probably don’t understand. that’s not to say i understand it either, but with the recent success of vine and the popularity and ease of youtube, i think we should expect more live-video user-generated content apps (like daily booth, for instance) to sprout up.
  • linkedin as a platform: we are all on linkedin and monitor the activity, but probably very few of us spend lots of time there. students don’t either, but they look at linkedin in a slightly different way — they see linkedin as a platform they can build new apps and services on top of. to date, companies in this category have tried to build slightly competing businesses to varying degrees of success, but the students i’ve seen want to create new things leveraging the network data locked in all of our work connections. i don’t know what the future looks like here, but i’d love to see someone bring me a product idea based on this.
  • founder drama begins early: perhaps this is just a reinforcement for us all, but one of the biggest risks in startups is co-founder risk. i can already see brilliant teams coming together to show me something or test a hypothesis, but the drive of competitiveness can sometimes also push back the hard, human work of trying to form and become a “team.” people use the word “team” as a given, but there’s a reason parents want their kids to play sports or other organized group activities — it instills the spirit of team that helps later in life in other pursuits, and no more so than company founding. having the (at times difficult) discussion around topics like equity splits is key to do early on.
  • scaling matters: a tricky situation for me is when a student proposes a great idea for a product or company, but when we dig into the plans, the initial team requirements pose an issue. salaries (and rents) are the two biggest things investment dollars are allocated toward, so investors who are like me are very sensitive to hiring projections because these impact burn rates the most. many of the students i’m meeting will be great ceo’s one day, but they also haven’t run a business yet, had to make payroll, had to fire someone, and all those hard things. they will, i’m certain they will soon, but they aren’t oftentimes exposed to these P&L realities in the classroom.
  • it’s not a race: at a place like stanford, right now it’s incredibly competitive. yet, at the same time, it’s not a race. there’s really plenty of time. it’s ok for some students to go into big successful tech companies, or to join a startup, or to go into vc — the real motivation should be to go to a place where they can continue to learn and drive their passions deeper. it’s hard for me to convey this because i respect and feed off of their enthusiasm, and while there are many technical and other problems in our world that need creative solutions, everyone will move on their paths at different speeds — that’s ok, and i hope people gain the maturity to make the decisions that are best for them as individuals.
  • our immigration policy is lame: i‘m not one to jump into political debates, but i do know this — when really talented engineering students from other countries approach me and are concerned about staying here, i can literally see the concern and dread in their eyes. they want to stay here and work. they are more than fully able. and, we have invested our own resources in continuing their education. forcing them to go back home against their will is a shame.

investment stakes


after about one year on sand hill road, there are a series of learnings that i would like to share with founders. 2012 has been a terrific, fun year. the valley and platforms are changing really quickly. at times, it feels hard to keep up with it. everyday, i get to meet people who are trying to do new things, or trying to make subtle improvements on existing things. it’s all really inspiring, i’m very lucky, and i’m having a blast.

and now, as 2012 closes, and i’ve led a few investments at kp now, i wanted to write this post to share a bit more about what my investment style is. of course, it can change over time, but knowing me, i’m sure it will stay roughly the same. people often ask me: “what do you like to invest in?” usually, that’s a hard question to answer, given my range of interests. and yet, i find myself thinking about it over and over again as the year ends, perhaps because i may finally have an answer.

here’s how i’d describe myself as an investor: “i’m a stage agnostic investor. for the right person (usually a former colleague), i’ll make a seed investment. for a later stage deal, i’ll get involved in a growth investment, especially if i can help the company or team scale faster. and, of course, my natural preference is to play around the edges of series A, because this is where i like to apply leverage, lead the round, and help the founders cross the first big divide.”

i’ll briefly explain this, as i think it will better help entrepreneurs understand what i’m looking for.

in series a, i like to lead. by nature, i want to find investments and founders where i can effectively lead a deal, participate on the board in some way, and really put my muscle into recruiting, technology development, and other key considerations. this means that you’ve already been working on something for a while, perhaps taken seed funding to begin with, and now are looking for a real investment partner. i’m lucky to be at kp, with a wealth of resources the firm as to offer, as well. basically, i want to find situations where my presence and kp’s presence could lead to outsized success and possible breakthroughs.

there are exceptions, always, right? at the seed stage, i will definitely consider outstanding individuals, especially those i have worked with and/or managed or recruited in the past. this is what i did with apcera, led by a brilliant technologist who i am privileged to support in this small and special way. i’m always happy to meet folks just starting out, especially if there’s a connection between us, but generally, i’m more likely to intro you to seed investors i respect that may fit. of course, i’m an email away, and if you’re building something great, chances are i’ll find you, too :)


this is who i am. it’s pretty simple. looking back a year ago, when i wrote a letter to my future self, i think largely i’ve kept true to that ethos. i hope so. and, from an investment point of view, whether it was gumroad or more recently, clearstory data, (and one coming soon, stay tuned), i find myself leaning in most at the series a investment level, and generally the question is simple — does it make sense for me to join this board, and would the founder benefit from it? in the rare cases where the answer is yes on both sides, i lean in, bring out the pen, and partner up to go to war. that’s the only way i know.

announcing the investment into clearstory data

i’m proud to announce kp’s investment in an exciting young company, clearstory data. you may have read about it last week, clearstory is an innovative company with seasoned people at the intersection of some pretty massive trends, so we are thrilled to be involved. i am leading this series A investment for kp and will join the clearstory board — and am also excited to work with investing partners andreessen horowitz and google ventures.

it’s no secret data is becoming more important. it’s past cliche, right? data impacts the large social networks. it impacts ad networks on the web and mobile. and, it even impacts governments and political power. we all know this. but if we focus on the realm of software for a minute (and there are many great data startups also working on hardware, servers, storage, etc.), the software opportunities in big data, especially at the infrastructure and application layers, are massive. with clearstory, the focus is on business analytics — itself a big category within data software applications — and fits alongside other great startups in this space, such as domo, platfora, and kaggle, among others.

the great coverage of clearstory data’s launch and announcement go into the product details quite well. on top of this, i’d like to personally highlight one dimension that excited me as an investor in the space. in addition to backing a great, small, lean, seasoned team with lots of experience in data, i’ve seen over the years, from my time at composite software to palm to twitter, and even at kp, that companies desperately want to leverage their data yesterday, but there aren’t many people with the requisite data science and computer science skills to help them. as a result, data functions and the data itself becomes put in a silo, accessible only to a few people. and, the most sophisticated tools in the space come at a cost, a cost that some larger companies aren’t as willing to incur. add to this security issues of larger companies, and most platforms run the risk of being stuck.

the reason i’m interested in a company like clearstory is because they’ve built out a product and modeled it to allow companies to simultaneously secure their data while allowing their people to leverage it with dead-simple interfaces. think of all the inefficiencies in modern work. the constant reporting, or seeing troubling trends too late. the data doesn’t lie, and a company like clearstory is on a mission to make data transparent, accessible, and actionable in today’s work environment. and, for me, that’s something i want to be involved with.

design in the kleiner perkins family

as you may have caught last week, kleiner announced a new program around supporting design as a craft. we all worked really hard to put this together and are grateful for the support from our founders and former colleagues to make this happen. and, we were lucky to get some great coverage for it too. in a nutshell, as we learned from our engineering fellows program, we wanted to devise an infrastructure where we could help emerging designers get great, hands-on experience and match them with designers who are already practicing in their field. the result is the creation of a design fellows program, which designers can apply to, and a design council, which will help mentor the next generation of great interface creators.

i also wanted to share, from an engineer’s perspective, just how critical design is which I fully appreciated during our development of webOS. everyone knows it takes great engineers and designers to create great products. twenty years ago, engineers had to create all UI, as not many others knew how to build in MFC. ten years ago, designers were doing production work, but interaction design was really nonexistent. fast-forward to today, and a design team is focused on human interaction, including user research, visual design, and interaction design. a few years ago when i was at palm, the webOS design team worked with the development team through a process to ship v1.0 and 11 post-releases (shout out to matias duarte). as we were leading these discussions, we fielded ideas from both designers and engineers, but at the end of the day, designers won the vote (pending the ability for engineering to execute, of course).

but as products get more complicated and have more usage, product managers emerge to handle product management, communication, and advocate for the customer. the problem here is that having too many PMs in the kitchen fighting over turf, credit, and features, the engineering team can get thrown in all sorts of directions. it’s not that different from politics, in some cases! and in the case of engineers, they oftentimes are called on to build *and* maintain, the latter which is often greatly underestimated and can be a thankless job in and of itself. more on this topic in a later post.

i’m not an expert in organizational structures, so i’m only speaking here from first-hand experience from company building, intuition, and gut. for each product or company, there’s an optimal number/ratio of engineers, designers, and PMs…all play a crucial role. at kleiner, and looking ahead to how these structures evolve, especially as phones and tablets become customers’ primary computing devices, the importance of design will — we believe — only grow. this is not to say that engineering or product management will recede, but rather to acknowledge that building strong, deep capabilities and networks in this artful skill will be a great benefit to everyone in the kleiner family.

the next evolution of the designer/engineer founding team may be the addition of a domain expert which would culminate in cross-disciplinary innovation –  but lets discuss that in a future post.

founding my first company

it was fall of 2001. i had just wrapped up selling electron economy to viewlocity, which has an intense experience of growing and then shrinking a team. during this process, i was the sole executive driving the acquisition and dealing with major investors. once the ink dried, i thought about taking a break, but i didn’t. instead, i holed up in my house and got wired in.

the fall of 2001 was a relatively quiet time. i mainly stayed at home and began building what would become the first real company i would found. at the time, i didn’t think of myself as a “founder.” i was just building software or talking to pilot customers. zero distractions beyond that. i financed the company for the first nine months, recruited a small core of engineers to the team, and we all worked out of my house. pure awesome times.

it was crazy and fun, looking back now. those were great times. it felt like i was really living in silicon valley (but then again i am a 4th generation bay area native!). eventually in august 2002, we raised a $5.2m series a. i moved the company out of my house and became the ceo of the entity: composite software (original name -> codemetamorphosis).

the idea for composite came from a combination of trends. at electron economy, i began to see that interfaces to data were becoming more standardized (with soap, etc.), and historically many of the challenges with data integration had been addressed with adaptors. companies such as crossroads software and active software were started to address this basic issue for enterprise app integration. these forces, plus some of my previous work in bioinformatics (querying multiple public genomic databases with local private databases) provided a background in building federated query engines.

the vision for the new company — composite software — was to build a data virtualization layer whereby the data could stay in the original system and our software could build a queryable composite view. large data lived in multiple heterogeneous silos, and getting a single view of an entity (in our initial case, a customer/client) is a very hard problem, both from a technical standpoint (high-performance queries on vertically partitioned data) and an organizational perspective (in their context).

in the fall of 2002, a year after incorporation, composite was lucky to have a few customers and seven-digit revenue. over a five year journey, our founding team grew the company to where it is today, still alive and doing well, with over 100 employees and many diverse customers and partners. these days, i advise the ceo. i recruited him into the company early, as well as several of the original engineers that worked out of my house are still with the company a decade later.

growing up in the valley, the dream of building software has always been around me. electron gave me a taste. but founding composite was the definitional move in my career. even though electron flew high and then was grounded, many opportunities and connections came out of that experience, especially interfacing with new investors and management. but, something deep down told me turn inward and just write code/have fun. i didn’t want to manage again just yet. i was a bit tired and exhausted from all the busyness of business. in a way, it was easier and more comfortable to build a new piece of software and re-recruit some of my colleagues than to go out into the real world. and, it was one of the best (and most fun) decisions i’ve ever made.

now on the venture side, i am grateful to be in the position to evaluate all these great new people, technologies, and business models. it’s like a never-ending stream of innovation, all coming by the kp offices. it is humbling. and, yet, we can only invest in a few teams each year. we have to be really selective. we have to consider the market size. we have to consider the competition. those are all the harsh realities of today’s technology market.

yet, at the same time, for better or worse, i am trying to find someone who wants to go through what i did. someone who wants to recruit teams but also has the moxie to know when things aren’t working out. someone who has already been working out of their house. someone who has convinced their former team members to go into battle with them. it’s not because this is the only path, or the best path — there are many paths. but, this is the only path i know, and if you’re not on this path, there’s a good chance i won’t be able to help you in any meaningful way. in future posts, i will share many of the learnings that i made throughout founding/building composite software. stay tuned.

kp in sf

in the world of software and the west coast, it sure does feel that there’s been a significant shift underway, north to san francisco. i could feel it for myself, even before i started at twitter in 2010. being at twitter in the heart of sf it was such a wildly different experience than palm and page mail road, and though i’ve grown up and lived and worked in the heart of the valley for so many years, it was clear that something new was swirling around the city proper. nearly the next two years were spent taking the 280 into twitter hq daily, a new habit that became a ritual.

now after a year at kleiner perkins, my colleagues and i have been observing this trend, as well. it’s not hard to spot — just glance over our calendars. while the center of gravity for kp is and will always be on sand hill road, the firm did secure an office in the city about a year ago, and we operate out of it while we are in the city. sometimes i’m there, jordan is there often, and so are chi-hua, megan, ray and bing when he comes up. plus so many others.

i’m in sf a few days a week, but i’m still working south of sand hill road quite often, and that area is still — and will always be — critical to the heartbeat of the valley. i’m excited to continue to spend more time in sf and to reconnect with old twitter friends and meet new companies. just drop me a note if you’d like to come by some time, and you may just see me walking around soma anyway. also, if you have any ideas on what you’d like to see us do with our sf office, please let me know. it’s a quaint space, not really that big, and not something that warrants a big press event, but we do want to create another space where friends in the ecosystem can come to and meet others, as well. i’m looking forward to your ideas, and thanks for reading. see you soon, either on sand hill or in the city!

announcing a new kind of technology show

i have an exciting announcement to make, and one that may surprise former and current colleagues…i’m really excited to announce that i’ll be hosting a new video interview series on techcrunch tv (TCTV) called “founders’ stories – engineering companies.” the concept is simple and straight–forward: on the first wednesday of every month at noon pst, TCTV will publish a short video conversation between me and a founder focused exclusively on some of the key technical challenges they faced in the creation of their products and businesses.

i’m personally excited for this new opportunity because it gives me yet another excuse to do what i love most — talk directly with technical founders and learn to understand the inner workings of the technical issues they’ve faced. the beauty of making it into a show is that these conversations become “open-sourced,” in a way, and will hopefully help transfer and spread knowledge within the community through a conversational format. (another potential win in this series is that future founders may better understand how a technical former-operator like me may be thinking in a venture context.)

enough about me, though. it’s all about the guests, and we have a great lineup on tap, such as bradford cross from prismatic (we’ll discuss balancing design and engineering decisions and other topics), a number of founders both at the later stages of their company’s evolution and others that are very early, and more. of course, i’m open to suggestions for guests, so feel free to contact me.

thanks in advance, also, to all the folks at techcrunch and TCTV for the opportunity. if you have any suggestions for guests, or topics to drill into, please don’t hesitate to comment or shoot me a note

here is the first session: