Who I am? Technology Innovation - I know something about lots of things (and most things about a few things)
Draft version, incomplete (TODO Also want to add leadership, business/CTO skills).
"Skills mixing a key to innovation"
This recent article in The Australian hightlights that technology innovation requires a mixture of skills, and that Australia has a problem:
"What we found was that they thoroughly mix STEM — most typically technical skills — with a range of non-technical skills broadly derived from the humanities, arts and social sciences (HASS), to meet the challenges thrown up by digital disruption, the volatility and increasing global nature of markets, and the war for talent.
...
The Global Innovation Index reveals that Australia is a relatively inefficient innovator. Australia’s overall ranking for innovation inputs is reasonable (10th). However, the overall ranking for innovation outputs is significantly lower (24th). The index indicates that Australia generally has the relevant skills but lacks the capacity to manage and use these skills and other inputs for innovation.
...
This predicament is in marked contrast to the experience reported among highly innovative organisations profiled in the ACOLA report. They recruit and develop employees with broad knowledge bases and strong integrative skills (beyond a single discipline). The bundles of required skills vary across the innovation cycle and include technical skills (science, technology, engineering, mathematical, digital) and non-technical business skills (business, management, financial, marketing) as well as creativity, design, interpersonal and entrepreneurial skills."
From their vantage at the apex of the innovation system, these enterprises tell us that STEM inputs are increasingly necessary, but not sufficient, to turn technical skills and knowledge into innovative outputs.
And a similar message from Wikipidia:
Innovation skills are practically the types of skills that allow individuals to become innovative in what they do. These are usually a combination of cognitive skills (e.g. the ability to think creatively and critically), behavioural skills (e.g. the ability to solve problems, to manage risk), functional skills (e.g. basic skills such as writing, reading and numeracy) and technical skills (e.g. research techniques, project management, or IT engineering).
Another article I read today "Expert Culture has killed the Innovator in Workplaces" (19 May) has a similar message, that generalists are essential for innovation.
"New research also suggests the usefulness of having generalists on innovation teams, rather than specialised experts. Breakthroughs arise when someone can combine many ideas together.
The key to being one of these generalists is to think broadly (not deeply) about a variety of topics, rather than a single topic. As Bertrand Russel once wrote:
There is something lost, when many authors co-operate. If there is any unity to be gained in a particular idea, there is a necessity to synthesise all the relevant information "in a single mind".
The solution to our lost innovative capacity might therefore be relatively simple.
...
Funding and grants should also be less tied to how much someone has specialised, and more to how inventive or creative they actually are. Our system should move more towards this kind of merit-based ideal, rather than focusing on expertise and certification."
Oh, and this blog is all about my "journey" to yet more specialisation: AWS certification. Perhaps I should ditch it and go back up a level to generalisation again?!
Creative Juices
During his working life my father was a keen photographer (movies in 8mm, and still photography, for example for documenting some of his civil engineering projects) and after he retired he took up painting (landscapes mainly). I have some of his genes and have always been keen on photography (particularly outdoors, travel, etc) to. I used to know the best film for the light and subject, and briefly had my own darkroom (black and white only). I did candid wedding photography for a couple of years (not the formal stuff). I've used a Sony and more recently LG smartphones to ensure that I always have a DLSR quality camera on hand as you never know when a photographic moment can happen. Over the years I've built up a reasonable collection of art and photography books (techniques and photographers works).
I've also dabbled in macro- and micro-photography (a rugged pentax bushwalking camera I bought a few years ago for the Routeburn track in NZ had really impressive macro and micro (1cm focal distance) modes! WG-2?). More recently I did an advanced digital photography and printing course in Canberra (over 3 months) which included theory, labs (photoshop) and practical exercises, with photoaccess. I did a project photographing the "nut guy" at the apic food markets. I took candid photos at my son and daughter in laws wedding in 2016. I got a nice "lucky" shot after all the formalities were over, they were back home sitting down on the grass in the back garden when the pet dog and 2 rabbits ran over and plonked themselves down on their laps. Couldn't have been planned if we'd tried.
I've also dabbled in macro- and micro-photography (a rugged pentax bushwalking camera I bought a few years ago for the Routeburn track in NZ had really impressive macro and micro (1cm focal distance) modes! WG-2?). More recently I did an advanced digital photography and printing course in Canberra (over 3 months) which included theory, labs (photoshop) and practical exercises, with photoaccess. I did a project photographing the "nut guy" at the apic food markets. I took candid photos at my son and daughter in laws wedding in 2016. I got a nice "lucky" shot after all the formalities were over, they were back home sitting down on the grass in the back garden when the pet dog and 2 rabbits ran over and plonked themselves down on their laps. Couldn't have been planned if we'd tried.
I used to be musical and played the electric guitar in a couple of pop-up bands while at university. I liked electronic music as a kid, and built (and tried playing briefly) a Moog synthesiser. I also had a Yamaha DX7 in the 1980s (1st commercially successful digital synth), which was one of the coolest synthesisers of all time, I played it (a bit) and programmed it (a lot). I also built a computer with a friend so we could experiment with electronic music while at university.
I (tried) playing a bit of Jazz guitar with small groups at UNSW colleges (electric guitar + clarinet, trumpet, trombone, bass, recorder, bongos, piano) in the 2nd half of the 80's.
I like cooking, but it's more "experimental" (creative). I like trying out new recipes (or ideas) and cooking things for the first time and trying them out on people.
I'm creative with ideas, I like to get lots of stimulating ideas and influences in a rush, think about them for a while, bounce ideas off people, and then come up with lots of new ideas and directions and to synthesise them together (possibly requiring further research). This process tends to be iterative and ongoing (with stops and starts).
I like art, history and technology appreciation, I particularly enjoy visiting art galleries, and exhibits, museums, historical locations, castles and forts, new interesting and large cities, and locations of natural and wild beauty.
My wife has been studying a fine arts degree (in sculpture) at Sydney University for the last 4 years. I've helped out with some of the more technical lighting problems a couple of times (e.g. solar powered lighting effects, and music activated lights effects, for 2 outdoor art installations).
I've been a judge for Tournament of the Minds, which combines design , technology and problem solving.
I've like going out to watch plays (modern and historical), and watching some modern dance.
I've invented and prototyped a few computer games (mainly the concept and the engines). e.g. a game world that worked on "magic", it was a simulated based world run by a Forth-like language. The world and player interaction used the same mechanism, and the goal was to learn the "magic" language (the Forth-like programming language) when enabled you to take more and more control over the world, set up automatic events, and eventually being able to manipulate other objects and players and even changing the underlying "physical laws". I've also played with variants of traditional board games (e.g. a version of Risk where the 1st phase is to create the board layout, and which allowed "pirates" who have no permanent land bases, and a version of Cluedo that worked with different historical Stately Home plans. E.g. The butler was killed in the drawing room of Buckingham palace with a corgi dog bowl!?).
I like reading and watching science fiction (and have a largish sci-fi book collection). Science fiction is all about creativity. What if technology is different, what if the phsyical laws of the universe are different, what if society is different, etc. They are essentially thought experiments.
Sci-fi movies? Lots, 2001: A Space Odyssey (still have the sound track on LP), 20,000 Leagues Under the Sea (the original movie), the 1st movie I recall seeing with my father in the theatre in Auckland that had stars all over the ceiling, magical, I was about 5 or 6 at the time. It wasn't my imagination, it's the Civic: "The interior of The Civic is famous for its domed ceiling of stars. It is, in fact, a flat ceiling cleverly painted to resemble a dome, with thousands of lights as the 'stars’. The 'night sky' is a representation of a Southern Hemisphere sky at 10.00pm on an April evening":
The dome of the Civic in Auckland.
And my all time favourite sci-fi movie, The Adventures of Buckaroo Banzai Across the 8th Dimension! What happened to the sequel??? The credits mention a sequel, Buckaroo Banzai Against the World Crime League, which was never produced. Very Sad.
Entering (or possibly exiting) the 8th dimension!
Is reading creative? Not sure, the authors are certainly creative geniuses however, and I keep hoping that some of it will one day rub off on me.
Some of my conference presentations have introduced a creative/demonstration element.
For example, in Heidelberg (Germany) for the Middleware 2001 conference (Paul Brebner and Shuping Ran, “Entity Bean A, B, C's: Enterprise Java Beans Commit Options and Caching”, Proceedings of IFIP/ACM International Conference on Distributed Systems Platforms, Heidelberg, Germany, November 2001, LNCS 2218, Springer-Verlag, pp 36-55.) I used balloons to demonstrate cache hits and misses. Popping a balloon was a miss! I recall someone had been "sleeping" and woke up with a start when I popped the 1st balloon!
For this presentation at the Borland Conference (Paul Brebner, Is your AppServer being crippled by the JVM?, Invited talk at BorCon 2002, in proceedings of the 5th Annual Borland Conference Asia Pacific, 2002, Sydney), I used a rubber chicken and eggs (some real, some blown out) to demonstrate the impact of object lifetimes and JVM settings on AppServer J2EE performance. I "accidentally" broke some of the eggs to prove they were real, and then threw some out into the audience during the talk (I think everyone managed to catch them as they assumed they were real - the ones I threw out weren't :-)
I've a also used "simulation" role-playing games (using people as the actors and providing short written instructions) for a number of talks, including sensor network architecture (with increasing arrival rates), and a software engineering simulation game (high school, to learn about software engineering for an imaginary games programming company) and simulating the internet! Lots of fun.
Startup Experiences
I worked for about 5 years in a tech startup in the 1990s (Softway Pty Ltd, UNIX systems programming and resource management product) which required a change of mindset and skills from pure R&D to more client focussed engagement. We had a mix of small (consulting) clients and a few larger R&D contract clients (e.g. NASA, SGI, Fujitsu). The larger clients were more challenging and required more long term engagement and attention to detail and quality (e.g. documentation, bug detection and continuous process improvement). Some of them also required us to take into account tricky cultural and business differences (e.g. Fujitsu). I also became proficient at working with smaller clients and finding creative ways around their sometimes arcane communication and business processes. Agility was often key and using technology to our advantage (e.g. model driven tools so we didn't have do rework all the time when the specifications were changed on us at the last minute, we just changed the model and regenerated the code). We also did something similar for larger clients. E.g. the tests were automated, could be run on different platforms and take into account platform specific features, and dynamic generate and execute the test-cases. We had a company "suit" that we could use in an emergency if we had to meet with "suit wearing clients" if we were't prepared in advance.I joined NICTA as a senior research scientist in 2007 in the e Government project in Canberra. I was responsible for "Use" inspired research and development driven by real problems in large scale software projects. This was run like a startup in that we had real paying customers with practical (and messy) problems from early on. This enabled us to grow out team. We had to develop solutions that worked on real world problems (e.g. imperfect information, changing environment) with real world constraints (e.g. time bound, immature organisations, etc). We had to make our software more usable, quicker to produce results, and to produce simplified targeted results (to answer specific business questions). All while doing ongoing R&D making it more complex internally to cope with more demanding problems. This was good practice for juggling client engagements (sales, contracts, legal, IP, project planning and management, customer relationship, and the actual technical work and delivery of results), software project management and architecture, and ongoing research (e.g. keeping up with technology trends, communicating our work at conferences including professional and business venues, etc).
About 5 years into this work NICTA was keen to spin out more startups and our focus on clients and having a working tool appeared to be a suitable candidate. I was involved in the spin out process for a few years which included management of IP (lots of meetings with IP lawyers), spinout contractual negotiation, some training around finances, psychology, pitches and selling, etc, and meetings with NICTA executives and the NICTA holding spinout company and NICTA board and advisory committee members. Leading up to the pinout I wrote a patent application in a week which was formatted by the IP lawyers and submitted just in time. See also thoughts on being involved in a tech startup.
CTO/Chief scientist/director
TODO Just a quick summary at present
Role as CTO has included being Chief Scientist, Chief technical consultant, Director, Product Manager and Architect.
Chief scientist: About 1/3 of time. R&D based on new client requirements, research of new algorithms rapid prototyping, evaluation of 3rd party/open source tools, simplification of GUI and how results are presented, rapid development of new pre-processing pipelines/integration to client/vendor products (e.g. Compass, Dynatrace, AppDynamics).
Chief consultant 1/3 of time: Pre and post sales with clients, contract negotiation and pricing, legal aspects, dealing with security requirements for data movement, SoW development, project management when required, customer relationship management (mainly technical side), initial and follow up meetings for projects, obtaining data (including architectural, test, APM data etc), initial analysis of data for quality and quantity, initial model building and predictions, iterative refinement, dealing with any issues that arise including product enhancements, data problems, increased or changed scope of what's expected, time problems, etc. Authoring and presenting interim and final reports etc.
Explaining what the results mean in terms of impact and how to action them (particularly tricky when results have error bounds which they always do).
Product manager and architect 1/3 time: Overall product management and roadmaps from technical and user requirements perspective. Introducing and retiring new languages and frameworks. Longer terms planning for product enhancements (e.g. adding tags to components, new features requiring different and more data from APM tools, automatic sensitivity analysis and error bounds, new approaches based on generative modelling and transformational approaches rather than model/simulation as current, etc)
Business/Director 1/3 time: Director of company, financial and strategic oversight, business and prospect planning and monitoring, end of year reporting and accounting, technical risk assessments, finding new avenues for getting work (e.g. DTO services panel), putting in proposal for Government grants, IP management (Patent renewal) and protection of code, Communicating to professional and academic conferences and vendor relationships (e.g. forums, partnership and technical development with Dynatrace, etc).
And yes it does add up to 4/3 :-)
Other leadership experience (non-profit boards)
TODO
Business Skills
While I was a visiting research fellow in Computer Science at UCL in 2004 I read the preparatory books (e.g. "The 10-Day MBA"), and based on submission of a cv and a project proposal I was invited to enrol in the London School of Economics Technology Innovation programme for practicing professionals. Unfortunately the timing didn't work out as the course extended beyond the end of the year (when I had to return to Australia).A few specific financial skills I picked up over the years on various non-profit boards and as CTO for the most recent startup are::
Financial Trend Analysis
For some types of companies income is relatively constant and predicatable. This was mostly true with the non-profits I was on the board for, but not always. So you could do financial projects based on income this year compared to budget and previous years income/budgets. However, there was some "down" times of year and other hiccups (e.g. after creating a "spinout" child organisation which took 100 of your voluntary donors with it). In cases like this I found that using regression/trend analysis over longer time periods invaluable. If the general trend was down (within error margins) then there really was cause for concern and you had a few months to panic and attempt to do something about it.Prospect Probability Analysis
If you have a number of potential business leads or prospects in your pipeline that you are pursuing, all with various likelihoods of eventuating (guesses really) and guesses for how much they may be worth and when (so three variables, probability, value, timing), how do you know what the probability is that you will still be in business (and by how much) next month? I.e. what is the expected value of new work that you are likely to have over time.It turns out you can't just add the probabilities together (really? Who would think that?). I.e. if you have: prospect A with 30% probability, prospect B with 40% probability, and prospect C with 50% probability, you DON'T have a 120% probability of having work as they are independent events. It's a bit more complicated than that sorry. There is theory. Good luck. I'm not sure any of the SaaS propects/pipeline management tools automate this however. Actually, the maths I used looks to be simpler than the academic Prospect Theory, possibly I invented a simplified version resulting from making more assumptions. An even simpler example is as follows. Assuming you have 4 prospects, and each prospect is only either binary (unlikely or likely, no/yes), then you can analyse it as for coin tosses. There is therefore 0.0625 probability of getting no work or all the work, and 0.9375 probability of getting at least some work.
| Number of Heads | Number of Ways | Probability |
|---|---|---|
| 0 | 1 | 1/16 = 0.0625 |
| 1 | 4 | 4/16 = 0.25 |
| 2 | 6 | 6/16 = 0.375 |
| 3 | 4 | 4/16 = 0.25 |
| 4 | 1 | 1/16 = 0.0625 |
However, with more probabilities things are less optimistic (similar to throwing a 10 sided dice).
For example, some results as follows. If you have 5 prospects with 0.1 probability each (1st 5 rows in the table below), you have 0.41 probability of 1 or more sales, but 0.59 probability of NO SALES. If you have prospects with increasing probabilities from 0.1 to 0.5 (2nd 5 rows), then you have 0.84 probability of 1 or more sale, and 0.16 probability of no sales. If prospects have likely income of 20, 50, 100, 200, and 300 respectively ($1k if you like), then the average expected value is 67 for the first set of probabilities, and 272 for the 2nd set. However, averages are not all that meaningful. Better is to use percentiles. So, what's the value expected with >= 80% probability? For the 1st set, the 80% expected value is actually 0. Oh. For the 2nd set, the 80% expected value is 50. That's not very high either. And if the prospect values are reversed (final 5 rows) then the 80% expected value is only 20. NB These results were computed using a simulation program with 1 million experiments.
You can also answer questions such as what's the probability of earning more than $100k next month? For each of these 3 scenarios the probabilities are 0.27, 0.78 and 0.49.
Here's the same data in table form.
You can also answer questions such as what's the probability of earning more than $100k next month? For each of these 3 scenarios the probabilities are 0.27, 0.78 and 0.49.
Here's the same data in table form.
Want to change your career in Selenium? Red Prism Group is one of the best training coaching for Selenium in Noida. Now start your career for Selenium Automation with Red Prism Group. Join training institute for selenium in noida.
ReplyDelete