Electricity transformed countless industries.
How are you Plugging In?
A Low Risk Entry Point to AI
Week 1: Strategy and Idealization
We’ll identify and refine your key value proposition, project objectives, verify it against the state-of-the-art AI technology and design patterns. We’ll apply the most effective combination of learning algorithms to the data and business problem that you are most eager to solve. We’ll provide you with a roadmap that is customized to your business needs and highlights specific opportunities were AI can be leveraged.
Week 2 - Deep Dive and Roadmap
From evaluating data sources, ingesting structured and unstructured data, data munging, to creating training sets that jump start the Deep Learning process. By week’s end you have a comprehensive end-to-end roadmap you can take to any Data Engineering team to get you going on the Proof of Concept – Pilot Project – Production Environment journey.
Why 'Research Machine'?
The reason we exist is to accelerate the actualization of Deep Learning instances in the real world. Read: we want to cut through the clutter of current academic journals and newspaper articles, and get this AI in the hands of practitioners. The broad field of “Research” is being affected by many forces: the routine of updating excel rows and columns with even more data, the explosion of more fields to fill in, the explosion of unstructured data that can give more clarity to an analyst’s insights, the increase in frequency of this data availability – from quarterly to by-the-second. All these data and how they are analyzed support evidence-based investments and decision-making. It is clear that a faster and more comprehensive way of getting ahead of this explosion of opportunities is through Artificial Intelligence, more specifically Deep Learning. We created Research Machine to be the tool for research-intensive organizations.
What is Research Machine?
Research Machine is Intuition Machine’s suite of services that enable Deep Learning in a business workflow. It shape-shifts, depending on the chosen entry point into Deep Learning. It is a Roadmap Project Plan, which details the steps that a group needs to take to get started. The Roadmap leads to a Proof of Concept, which is a learning vehicle for tweaking what the next step, the Pilot Project, would look like. The Proof of Concept stage is also where a good fitting is made with ongoing workflows of the group. Finally, the Pilot Project gives way to a Production Environment, which is a point where the group begins to derive benefits from using Artificial Intelligence in the workplace.
Explain the Iteration and Learning Process
Sure. Just like a startup, the use of Artificial Intelligence is based on a number of hypotheses. “We can save on headcount expenses by using AI“, or “We can augment the work of our current analysts with AI“, or “AI can help our analysts absorb more, and more disparate information” or “We can gain more customers when we show that our work is augmented by AI algorithms.” All these hypotheses define a certain goal for using Deep Learning in the workflow, and therefore, determine what milestones to cross on the road to successfully proving the hypothesis. This will require iteration through many ideas for how to implement the Proof of Concept -to- Production Environment journey.
Are you a hardware or software company?
Yes. Ultimately, what enabled the stunning advances in Deep Learning is a combination of: very large data setsto train on, very fast hardware (GPU – graphics processing units) that can perform the algorithms required by the third component, the software patterns defined by the problem. You have a choice of hosting these data and software in the cloud, or on-prem, via hardware and software (which designs are mostly open source, thereby reducing cost of entry).
Who can use Research Machine, and why?
Research Machine if for teams, groups or businesses that analyze and report some conclusion based on research. This can be as large as the largest research departments on Wall Street, or as small as the independent Research Consultant or Equity Researcher. It can be the bond trader updating information on multiple portfolio elements, or the distressed securities analyst going through mounds of discovery documents. It can be the oil trading group analyzing changes in satellite imagery of Rotterdam oil tanks, or a retail analyst evaluating shopping mall parking lot congestion. Research is not just just for the investment and finance industry. All around the world, huge strategic decisions are being made and validated based on competitor analysis. Sensors, tweets, blog posts, customer comments on public forums, conference results – all define a large and growing source of data that must also enter into the research equation. We believe the Deep Learning software in Research Machine can help get ahead of this growing threat – and opportunity.