top of page
Image by Pierre Borthiry - Peiobty

dōnō analytics, research and knowledge (DARK) lab is a small lab from a small organization, which intends to do interesting work and extend, however marginally, the body of knowledge in areas of its interest.

Research agenda

There are three broad areas for the lab which is in our name. Within each area, there are specific domains where we have more expertise than others, but that does not mean that we will not collaborate in those as well.

  1. Analytics – we intend to work with analytics techniques to understand different types of problems which may or may not have been explored before. Our approach will be defined by our multidisciplinary background and interest. Both of these are a part of our research agenda.

  2. Research – we did not call ourselves an ‘applied’ research laboratory. Because, we do not want to stop our colleagues and collaborators from doing theoretical research. The focus of the lab is research and it may or may not have any applied or commercial value.

  3. Knowledge -  Our definition of knowledge goes beyond the academy – it includes not just academic knowledge but practical ones. We intend to create case studies and research papers, to disseminate knowledge that is getting created by businesses through their daily tasks, and we also intend to write white papers and popular press articles to disseminate knowledge which is often closely held by a somewhat ivory tower academic community.


How can you join the movement of small corporations research?

Theoretical research is for universities, applied research is for large corporations, and the rest try to develop products based on research but rarely spend on research. We would like to tell Mr. Friedman that corporations are smarter than that and shareholder value is not and should not be everything a corporation is about. Our small way of helping stakeholder interests is by doing research (even as a small, bootstrapped organization). If you are a small corporation who also does research or wants to do more do reach out. Co-opetition beats competition hands-down, every time.

What are some of the research projects we are working on?

  1. Real estate is core to our work with many clients and we are working on figuring out ways how models can help increase transparency in emerging real estate markets. While we have already spun out PropMath, we are working closely with that team on building some unique models of analyzing real estate.

  2. Are regular people rational when they lend money to strangers? Using millions of data points of accepted and rejected peer-to-peer loans, we are trying to understand the rationale behind loan giving. How smart are lenders? Are they biased? Even though they do not use data science, are they smart enough to make good decisions?

  3. Using machine learning algorithms to create better prosthetics. This is the PhD topic of one of our colleagues and we are extremely supportive of this exceptional idea for an incredible cause.

  4. Are their directional biases in forecasts from central banks or global institutions? We know forecasts are often wrong, but can the direction or degree of ‘wrongness’ be predicted? And if it can be predicted, does this lead to a tradable idea?

  5. We work and have worked extensively in quant finance. There are multiple areas of quantitative finance research where we are dipping our toes. At its core is figuring out relevant financial data and how to slice and dice and clean them to come up with sensible data to predict at small intervals.

  6. Application of non-linear dynamics is of great interest to us since we are bored of linear and tree based models J

  7. Econophyics makes some exceptional predictions and provides insights. But, econophysics is limited by its simulation assumptions and often codifying human/trader behavior to 2-3 kinds. We believe behavioral economics can help econophysics – and we are trying hard to create a center for behavioral econophysics. Let me know if this interests you.

  8. Traditional machine learning algorithms and neural networks are extremely limited and causal ML/AI (with optimal explanability) will change the way ML-AI is done. We have implemented Causal ML and we have some ideas to do some applied research.

  9. We believe, using neural networks to emulate human behavior is like using atoms to understand the trajectory of a falling object – you can do it, but the physics of larger objects (mechanics) is easier and more helpful for this problem. Similarly, human behavior (and ergo, intelligence), we believe, can be much better emulated by algorithms from quantitative psychology, than the physical substrates of neural signal processing. We are working on multiple papers in this area.

  10. We are interested in digital humanities, which we believe is already an important field of study. We would be glad to collaborate in this area with folks who are experts because we are not.


Why should you reach out to us?

Academics, scientists, researchers, citizen scientists, graduate students or professionals who are interested in research and would want to collaborate in any of our fields of interest. Please email and if your interest and ours match anywhere, we will get back to you.

Why should you follow us on social media?

As we know more, we will write more on each areas of our research. We will reach out via

bottom of page