AI Playbook for Startups and SMB

The AI Playbook for SMBs and Startups

Most businesses in the SMB and startup class, still struggle to grasp the broad definition of AI disciplines, what AI really is to their business, where to get started and what successful execution looks like for their businesses. Simplifying a focus area narrowing these concerns specifically to address your business pressures is helpful to keep team in line with long term business goals and strategies.

There are many AI Playbooks written by renown gurus such as Andrew Ng, meant for Fortune 500 organisations and they do contain some very useful guidance and tips for all businesses. Nevertheless, smaller ventures with limited funding, skills and resources may need to rework tactics to suit business.

Here we present steps to formulate a generic AI playbook meant for small and medium sized businesses, which includes stages to:

• internalise AI definition for the business; 
• outline and prioritise business problems that can be addressed with AI;  
• experiment with ready pre-trained model
• build required skills or internal structures;  
• prepare training and test data; 
• choosing a path between pre trained and custom built model for solutions, 
• and assessing overall impact to businesses performance. 

The generic AI Playbook should ideally incorporate other dimensions unique to your business, industry and commercial eco-system over time, for maximum results.

Define and internalise AI for your business...

AI itself is a broad subfield of computer science. It encompasses many branches of studies including rule based AI and decision trees; machine learning (regression, classification, neural networks/ deep learning, reinforcement learning, supervised and unsupervised learning); robotics; text and speech processing (NLP, NLU, NLG); vision (computer vision, image recognition) and several other applications that still have not left the laboratories. 

AI is a field where one or several of these approaches are used to create machines, agents and models that perceive, learn and adapt to perform various tasks and cognitive function similar to humans.  

The AI domain is practically exploding with hundreds and thousands of new approaches, methods and solution areas in transfer learning (learning from lesser data and smaller models), networks with memory (generalising AI, e.g. DeepMind’s differentiable neural computer), reinforcement learning, training simulations, hardware for training and inference that are rapidly advancing in the research, development and real world application stages. 

Communicate how AI should be perceived in your business....

As such, it's easy for businesses to get lost in an ocean of offerings that comes with vendor specific methods, best practices, strategies, tools and ready use cases. Charting and communicating how AI should be perceived within the business very early on, in your AI journey is crucial to ensure potent utility, informed expectations and performance. 

For instance should AI merely assist workers to perform better at tasks? Should AI considered as an avenue to outsource cognitive functions otherwise performed by human workers ? Or should AI be a form of cognitive technology to help expand ordinary worker's ability to explore meaningful solutions to business problems both creative and logical?

Setting the right internal tone on AI can guide alignment with work culture, strategies, high impact projects, investments, selection of tools and services that can drive AI values across the business.

Define business problems that can be solved with AI

The common misconception in this stage is, data science or business team alone is responsible to outline business problems solvable by AI. 

Every business unit can contribute to this stage including sales, marketing, services and tech departments. Imagine how it will affect your business if tech can reduce 40% of cooling needs in your datacenter or how your customer service or inbound sales department can improve performance by multi-fold with realtime speech to text analysis that prompts them with emotion, sentiment and other useful information during a call with a customer.

Build working team; conduct regular meet-ups; answer business questions around growth hacking, improving or staging customer experience, workforce productivity and innovation; keep a revisable record of issues pressing the business growth (this can be a simple excel sheet with list of items); prioritise problems to solve; assess data availability; experiment, assess skills requirement; select initiatives for executive sponsorship and create projects to execute. 

Assess data availability and prepare training data for the problem

Data sources...

Data can be gathered from within the business, public datasets, procured from external sources (e.g. Image Net, Google Data search) or a mix of these. Identify useful data assets for your business problem, list what data is missing or not available and which data can be excluded from your raw selections. Collect and aggregate sufficiently to fit your model. 

Data preparation is a significant contributor to cost of building AI solutions specifically ML models, aside from high performance computing resources though GPU innovations has brought so much economics since a decade ago.

Clean and transform data for training...

Most ML models requires large amounts of data to train and converge before they can be applied to business. Preparing these data may require domain experts to collect raw data, identify features with strongest prediction power and label them. Though the business problem you are aiming to solve, tools and selected algorithms ultimately influences data preparation approach and mechanisms. 

Data transformation is revisited several times....

It is common for developers to spend 50% or more time preparing and scrutinising data quality and revisit data transformation processes (or feature engineering) several times during the development of the model. 

Split data sets for training, test and validation...

The prepared data is usually split into training set to fit the model, test set to evaluates final model fit on the training dataset and validation set to evaluates model fit on the training dataset while tuning model parameters.

Consider pre-train models

Ready to train models for various use cases...

For common business problems which requires simple recommendations, classifications, regression, clustering, anomaly detection and ranking, look for pre-trained model offerings by public cloud vendors such as Google, AWS, Alibaba and Microsoft. Many large and niche AI vendors now offer pre trained models for other slightly complex solutions to perform predictions, speech processing, computer vision (video, image and object recognition), decision support and planning tasks. 

No need to build algorithm or ML models.....

These offerings saves businesses the hassle of building algorithms and models from scratch. Plus it's a great way to get started with applied AI for business with quicker results and fast ROI times without having to reinvent the entire process (the value is in the model outcome not building it, especially for non-tech businesses). 

No need to replicate work for  ML backend .....

The team need only concentrate on preparing data, select features and train the model for required tasks (e.g predictions, classifications).  Trained models/ functions can then be deployed with other production applications and systems through APIs provided by cloud vendors to apply learnings on new data samples without replicating any work for ML backend.

ML and DL frameworks

In the event that the business problem you aiming to solve is much complex and does not have a ready pre trained model available, it is advisable to custom build the model custom build the model on a machine learning and deep learning frameworknof your choice, offered by cloud vendors through their ML-as-a-service offerings.

Frameworks such as Tensorflow, Pytorch, MXNET, Keras, Caffe, Scikit, Microsoft Cognitive Toolkit and Theano, comes readily packaged with interface, libraries, tools, pre-built and optimised components to facilitates fast development of AI models without getting into the details of underlying algorithms and complex architectures. 

New developers can benefit from starting with frameworks such as Keras and delve deeper into Tensorflow or other suitable frameworks as they get accustomed with the building blocks and programming languages.

However, it is highly recommended for new teams to get experience from simpler projects involving pre trained models first before venturing into challenging custom building models. It is easy to loose ROI when tinkering with ML or DL models without past experience and benchmarks.

Building AI skills 

When it comes to smaller businesses, building AI skills in-house is really a trial and error exercise to find the right blend of subject matter experts (or the minimal quo) to address AI problems and build solutions. This team's skill level is adjusted according to how they plan to address business problems, for instance with ready models or custom build models.

AI requires programming (e.g. R, Python, Lisp, Prolog, Scala, C, C++ and Java) data science (at least fundamental level) and business domain knowledge at the minimum to obtain some high level results. A strong foundation in mathematical fields such as probability, statistics, linear algebra, mathematical optimisation is necessary if your business wish to develop own algorithms or modify existing ones to fit specific goals and constraints.

Training existing workforce, hiring from other organisations and university grads are some of the common methods to build the right team.

Education sites such as Coursera, EDX, Machine Learning Mastery and many others provide neutral self learning content suitable for both technical and business audience. Content varies from fundamental to advanced level and comes with certification process that can be committed at one's own pace. 

Vendors such as AWS, GCP and Microsoft too have their own academies but syllabus may be skewed towards respective services. AWS and Microsoft are so far the best corporate academies I encountered and complies to many criteria mentioned in Andrew Ng's Fortune 500 AI Playbook in addition to easy to follow content and no fuss lab use (note that this is paid resources).

Assessing impact to business..

Once trained, tested and deployed, weigh the model's impact to your business performance. I use a no fuss method where I categorise what was increased, reduced, created and eliminated over a set period of time.

For instance:

• Did the model INCREASE overall customers value, revenue or profitability ? 
• Did the model REDUCE fraud risk, customer churn or cost ?
• Did the model CREATE new growth engines, data products or differentiators ? 
• Did the model ELIMINATE manual task, blind spots or redundancies ?

These answers are then scored and quantified for best alignment with overall strategic goals and objectives. The data is then collated for all AI projects to visualise and quantify the total impact to business financially (profit, revenue, growth rates, etc) and other areas such as competitiveness, reputation, and customer satisfaction. 

This exercise is helpful in two or three ways :

• to identify successful AI projects and approaches
• to build AI projects that can increase the value or performance of existing models
• find optimal opportunities to apply AI learnings to other functional areas 

There are many ways to measure impact, choose one that you are most comfortable with and put it to work. Use the discoveries to increase effectiveness of AI projects for your business while filling gaps in skill sets and access to useful data assets.

Conclusion - Ethics, Explainability and Human-Machine Interaction

In the end, AI systems are nothing but stacks of infrastructure and programs in various architectures that learns and executes continuously, depending on its utility. The machines aid human workers to predict outcomes, navigate a situation and decide on best action path with less mental drudgery while spanning solution options that was never possible before.

But that doesn't mean AI systems and models are perfect or free from self vested exploitation. Since the 2010s public concerns about racial and other biases in the use of AI for criminal sentencing decisions, creditworthiness and other forms of social credit systems (e.g. China) is driving demand for transparent AI systems and agents.

Explaining decisions of AI algorithms or models, especially those powered by deep learning or other complex neural networks is still a difficult task and is also known as the 'interpretability' problem.  Unlike decision trees and Bayesian networks that are more transparent to inspection, complex neural networks stack on one another and may involve millions of neurones processing towards the outcome.

As such, deploying sufficient governance protocols for data privacy and security; an architecture which facilitates transparency and tools to detect bias in algorithms is as critical as creating successful models or functions. Though in many cases engineering for more explainability and transparency may lead to significant accuracy trade offs. 

Making Sense of Public Cloud Investments

The public cloud infrastructure services (Iaas) market, according to Gartner's estimation is reaching $60 billion in 2019 and is growing at 27% YoY (fastest growing cloud services segment) from the previous year. Assuming Asia contributes 25% of the global share, turns the region into a critical geo for cloud vendors to battle for market leadership. 

However, cloud practitioners in Asia are still somewhat battling with what is considered as decade old cloud myths evolving around security, compliance, IT control and short term ROIs when conversing with regional customers riding on conventional structures and work cultures. This includes financial institutions, communication service providers, retailers, manufacturers and other asset intensive industries.

While these are all valid concerns, over the years premium cloud vendors have more than proven that their infrastructure and platform services (IaaS, PaaS) are superior to any single institutional  environment in terms of performance, security, reliability and overall economics to sustain just about any type of workloads. The public cloud values are now evolving rapidly with introduction of newer, advanced services in opensource databases, artificial intelligence, low code coding, and running mission critical workloads, enabling businesses to gain better control of operations as they cut waste, optimise performance, develop the much needed new capabilities and differentiators faster than ever. 

Open-source Database Services

Businesses have been dabbling with open-source databases (as well as open-source operating systems, libraries and tools) even before cloud computing services emerged and was aware of the potential to free organisations from growing commercial database licensing burdens. However, managing these open-source databases required additional training, operational and maintenance resources, apart from the unpredictable fixes and releases from the open source community. These issues limited the usage of open-source databases such as PosgreSQL to a small number of non critical, in-house development projects and applications in the past. 

This scenario changed in the recent years as cloud computing companies started to offer fully managed open-source database services such as PostgreSQL, MariaDB, MySQL and many others. Businesses start to increase adoption overnight by consuming cloud based open-source database services for existing and new applications while gaining new efficiencies, never possible before. The polyglot nature of cloud native development, where multiple types of databases and models are deployed in an application composed of micro-services and API calls only drives this trend further. 

Popularity of open-source databases according to DB Engine Ranking site, is about to converge with commercial databases in 2019 or in the early parts 2020. 

Artificial Intelligence, Machine Learning, Deep Learning

Explore pre-trained ML models

If new to artificial intelligence and its sub domains machine learning, deep learning and transfer learning, take advantage of the various pre-trained models offered by public cloud vendors such as Google, AWS and Microsoft to perform prediction, classifications, natural language processing (NLP), speech processing, computer vision (video, image and object recognition), decision support and planning tasks. 

Start by experimenting solutions to business problems while inflating returns on organisational data assets. This provides an opportunity for the business to quickly reap benefits from readily available models where the team need only prepare data, select features and train the model for required tasks (e.g predictions, classifications).  Trained models can then be deployed with other production applications and systems through APIs to apply learnings on new data samples.

Building custom models..

As the internal team gain experience and get familiarise with ML projects, commence a process to create step-by-step AI playbook to define and prioritise business problems that can be addressed with AI, data preparation, algorithms selection, and developing a model to make predictions. This exercise should involve critical members from business, data science and the engineering teams, ideally aligned to addressing growth challenges, resources optimisation, innovation and enhancing customer experience. 

Machine learning and deep learning frameworks such as Tensorflow, Pytorch, MXNET, Keras, Caffe, Scikit, Microsoft Cognitive Toolkit and Theano, that comes readily packaged with interface, libraries, tools, pre-built and optimised components facilitates fast development of AI models without getting into the details of underlying algorithms and complex architectures. 

Pick a cloud vendor for ML-as-a-service that best fit your needs...

Top public cloud vendors such as AWS, Microsoft Azure, Alibaba and Google provide support to most of the frameworks mentioned above for custom building models apart from their pre trained model offers for common business problems. New developers can benefit from starting with frameworks such as Keras and delve deeper into Tensorflow or other suitable frameworks as they get accustomed with the building blocks and programming languages.

Develop cloud native applications faster ...

Today a challenge sparks an idea, an idea turns into a prototype and a prototype transforms into a multi-platform first release in just matter of days and weeks. Aside from core infrastructure services, cloud vendors offer various advanced services ranging from devop, language SDKs, container orchestration, serverless computing, no to low code development platforms (both proprietary and open-source), APIs, code libraries to ready templates to help organisations accelerate application development arising from sudden business needs. 

Most post cloud applications are series of micro-services connected to one another via APIs leveraging languages, databases and code bases most suitable for the required function and the supporting technical environment.

No to Low Code Platforms

In addition, low code platforms such as Mendix, Outsystems, Google App Maker, Appian, and Microsoft Power App empowers traditional developers, IT professionals and to a certain degree the average business users to accelerate software delivery for business use cases made up of common features and components by simply utilising existing templates, forms, objects, drag and drop of prebuilt features.

Ensure performance of mission critical applications....

Learn which of your mission critical applications are suffering from intermittent or persistent performance issues affecting business processes critical to productivity, customer experience and revenue activities. Cloud vendors offer various services and pricing plans to optimise business applications from SAP, Oracle, Microsoft, IBM and others where data grows exponentially with transactions in time.

Exploring public cloud offers for the aforementioned type of applications can reveal credible areas of savings, operational simplification, performance improvement, speedy delivery of new business requirements, higher utilisation, improved security and compliance. Keeping a revisable record of requirements and working with cloud vendors to formulate cloud contracts to efficiently counter all planned and unplanned workloads can free internal resources for myriad of higher impact business activities.

Know values, learn constraints and strategise for long term...

Ultimately cloud values can vary significantly from business to business depending on the nature of operational model and the overall ecosystem. The four areas mentioned here are general enough for most sectors, though some businesses might benefit immensely from other emerging tech such as augmented reality, IOT, 3D or bigdata platform services.

In addition, identifying constraints that prohibit enterprise architectures from extracting full potential of public cloud eliminates aimless unproductive explorations. For instance legacy apps on rigid centralised architectures that may not gain much efficiencies when moved to the cloud.

Finally having a solid long term strategy, helps. For instance all new development and deployment to take the 'Cloud First' approach, or maintaining a hybrid cloud environment with only unpredictable workloads moving to public cloud or 'Cloud Only' approach where the internal teams build in-house brokerage capabilities to fully monitor its environments on multiple public clouds and perform various optimisation as needed.

An Interview with the founder of Nanotechnology World Network

This week on Let’s Kopi we feature yet another exclusive interview with a tech enthusiast, Marine Le Bouar, founder and CEO of Nanotechnology World. She shares her views, various initiatives and the community that she is shaping simply for her love and passion for the nanotechnology field in solving various environmental, social and economic constraints that we face today. Here is her response to some of our questions in how nano technology will define future of computing and if “Moorse Law” is really nearing its limits.

LK: Tell us a little bit about Nanotechnology World?

Marine: Nanotechnology World is a network of more than 80,000 individuals and organisations who are leading research, development, manufacturing and commercialisation of nanotechnology worldwide.

At the heart of the network is the Who's Who of nanotechnology space which includes members from academia, industry, government and investors. The platform connects individuals, companies and products. We already have more that 20,000 listings in the Who's Who and are planning in doubling that number within the next few months.

The Nanotechnology World Network is uniquely positioned in this niche industry as it has the power to reach major players, researchers and industrials with just a push of the button. For instance, several "Ecosystems Partners" with whom we work closely with, help create use cases and simplify adoption of nanotechnology in industries.

LK: Can you describe your role in your organisation? 

Marine: My role in the organisation is central as I work alone! On a daily basis, it involves reporting the news, list the new job opportunities and events, work on several media partnership agreements, answer dozens of emails, manage the several social media platforms and help investors get in touch with companies looking for investments. 

Marine Le Bouar, Founder of NWN

Currently, I am working towards creating a user generated content platform to ease and diversify the content, and enable our members to list their jobs, events and news. My main goal is to promote the importance of nanotechnology and the amazing possibilities it creates, but also to facilitate its adoption in the industry. I truly believe it has the power to change technology as we know it and that it will helps us solve some of the major issues we have as pollution, cancer, energy and etcetra.

LK: Tell us a little bit about how you got started in the nano tech world? 

Marine: I have always been a technology enthusiast. From satellite phones, to extremely complex electrical simulators, or serial bus for high-speed communications and real-time data transfer, I spent my career developing markets for high tech products.

I vividly remember the day I read an article about a new material that heals itself. It was the first time I heard about nanotechnology and I was flabbergasted. At the time, nothing much was happening outside the laboratories, but I new I wanted to spend the rest of my career working in that field. I instantaneously felt it was the future of all technologies. I opened a group in LinkedIn, called Nanotechnology World, because I found the articles I was reading so interesting that I wanted to share them with other technology enthusiasts. That's how it all begun.

LK: Lately there has been many talks on how the 'Moore's Law' is reaching a limit, what are your views on this?

Marine: This question clearly divides the science community! I am an eternal optimistic. I don't think Moore's Law is dead. With Intel’s release of a 10 nanometer chip in 2017, that will be cheaper than its predecessor, at this point I consider Moore's Law alive and kicking!

LK: How do you think the computing world will change in the future? (as sizes of clouds grow, emergence of neural networks, new data types, IOT, AI etc)

Marine: It’s a mystery why Moore’s law (it's not really a law, it is a prediction) still holds true after a half century later and the computational growth it predicts will continue to profoundly change our world. We’ve just seen the beginning of what computers are going to do for us:

In-memory computing 
Graphene-based microchips or Graphene — one molecule thick and more conductive than any other known material can be rolled up into tiny tubes or combined with other materials to move electrons faster, in less space, than even the smallest silicon transistor. This will extend Moore’s Law for microprocessors a few years longer.

Quantum computing 
Quantum computing uses quantum bits, or Qubits, which can be a zero, a one, both at once, or some point in between, all at the same time, opposed to conventional computer can only assign a one or a zero to each bit. Theoretically, a quantum computer will be able to solve highly complex problems, like analyzing genetic data or testing aircraft systems, millions of times faster than currently possible.

Molecular electronics 
Researchers at Sweden’s Lund University have used nanotechnology to build a “biocomputer” that can perform parallel calculations by moving multiple protein filaments simultaneously along nanoscopic artificial pathways. This biocomputer is faster than conventional electrical computers that operate sequentially, approximately 99 percent more energy-efficient, and cheaper than both conventional and quantum computers to produce and use. It’s also more likely to be commercialised sooner than quantum computing itself.

DNA data storage 
A little bit of DNA stores a whole lot of information. A group of researchers from the Swiss Federal Institute of Technology in Zurich speculate that about a teaspoon of DNA could hold all the data humans have generated to date, from the first cave drawings to yesterday’s Facebook status updates. It currently takes a lot of time and money, but gene editing may be the future of big data. 

Neuromorphic computing 
It's a computer that’s like the human brain—able to process and learn from data as quickly as the data is generated. So far, we’ve developed chips that train and execute neural networks for deep learning, and that’s a step in the right direction.

Passive Wi-fi
A new way to generate Wi-fi transmissions that use 10,000 times less power than the current battery-draining standard. While this isn’t technically an increase in computing power, it is an exponential increase in connectivity, which will enable other types of advances.

LK: How do you think nano technology will impact/ assist quantum computing breakthroughs?

Marine: Nanotechnology plays a major role in the development of quantum computing by creating new nano materials. 

For example, researchers from the London Centre for Nanotechnology at UCL have shown that the electrons in CuPc can remain in ‘superposition’ – an intrinsically quantum effect where the electron exists in two states at once - for surprisingly long times, showing this simple dye molecule has potential as a medium for quantum technologies.  

Also, the University of Maryland researchers have developed a method to quickly and inexpensively assemble diamond-based hybrid nanoparticles from the ground up in large quantities while avoiding many of the problems with current methods. These hybrid nanoparticles could speed the design of room-temperature qubits for quantum computers and create brighter dyes for biomedical imaging or highly sensitive magnetic and temperature sensors. 

Another example: the nuclei of Graphene Quantum Dots for quantum computers are nano crystals made of semiconducting materials small enough to exhibit quantum properties. Graphene quantum dots are ideal for use in quantum computers because they do not have a spin 98% of the time, greatly decreasing a tendency to interact with the spin of neighbouring atom’s nuclei that dismantles their undefined superposition state. 

While practical Quantum Computers have not yet been achieved, the creation of qubit control devices such as graphene quantum dots, ion traps, optical traps, and superconducting circuits allow computer scientists to continually improve the floating point operations per second (FLOPS) capability of Quantum Computers. 

Scientists at EPFL have now identified a new class of materials whose electronic properties can prove ideal for the implementation of spintronics. In a classical picture spin exists in either of two directions: "up" or "down", which can be described respectively as the clockwise or counter-clockwise rotation of the electron around its axis. However, the full picture is even more fascinating; the spin is a quantum property of the electron and can thus be in a superposition of up and down. Similar to the picture of Schrödingers cat being alive and dead at the same time. This makes a controllable spin state also a promising aspect for quantum computers. 

There are numerous examples showing that nanotechnology plays a central role in the development of quantum computing. It's such a vast and interesting subject. I really can't wait to see what the future holds, but I am confident, singularity is near.

LK: What advise you would give the technology community to stay relevant in the next five years?

Marine: Keep up to date with the latest breakthroughs. Invest in R & D. Join discussion groups and associations. Create a road map for your products that goes far in the future and do no hesitate to dream. It might seems like science fiction today, but could be totally feasible in the next few years.

Attending professional conferences and industry events puts you in the direct company of industry thought leaders. Regular lunch or coffee meetings with your mentors can provide you with more than just advice and support. Mentors can give you new insight and perspective into your shared industry.

LinkedIn groups are a highly under-utilised resource for accessing industry experts. Many people join groups but never read anything that’s shared. Members of industry groups often have discussions relating to up and coming trends. Following industry thought leaders on social media can provide you with quick tips, links to important articles, and new insights saving you precious time that you’d spend searching for the good stuff on your own. Look for your industry contributors on Twitter and LinkedIn. Join Associations and get involved!

I hope LK readers and followers found the insights shared here useful and enlightening. Please feel free to follow the Nanotechnology World group on LinkedIn, share your comments and views.