Previous:

Silver Hands Assist App

Next:

Learners Journey Analytics

3 thoughts on “Boettcher pitch competition submission

  1. Great idea to look to the crowd to scale a ‘taxing’, manual process of mapping tax rules to company processes to ensure compliancy in tax implementation. It would be extremely valuable if you can indeed establish a machine process only between NLP adding features and labels for the ML model (but I always see a role for a data scientist). I would love to see this fleshed out a bit with more context on who benefits the most from this, how this could be a platform between states and industry, and how it is maintained going forward.

  2. Thanks for the comments. (Hopefully, the additional info that got packaged as pdf comments came through).
    Immediate beneficiary is my company the purveyor the transaction tax engine, whose value is useful without the monthly data update that keeps the calculations accurate. Keeping track of changes in North America is laborious enough with the current process, executed monthly. With cloud-based deployments of our tax engine, customers are now demanding application of changes as they occur from jurisdictions, and the process is stressed (errors, lateness) as we try to do a couple of updates per month. Add to that the fact that there is plenty of additional content to add (verticals like tobacco, hospitality) in North America, let alone to support our aspirations internationally with VAT and Brazil, and we have a process that will not scale. Moreover, if a competitor figures our how to do this using analytics technologies (as a public company we know our current main competitor has purchased an AI boutique firm with the intention of pursuing it), our viability as a business could be disintermediated. Once in place, besides our use on “standard” content, we could make it available for customers to add content specific to their business/industry.

    As far as maintenance going forward, the thought is that so long as taxation approaches persist, a well-tuned analytics engine that picks up on changes and lands them into our calc engine’s format will hold up pretty well. As we add first time content for new industries or countries, there will be presumably be a training phase involved. Ideally this would not have to start from scratch with human researchers building tax rules, but rather could use the analytics engine to identify similar formulations in existing tax regimes.

    Serving as a platform between states and industry (i.e. taxpayers) is an interesting idea. We have often talked about having the states deposit their content directly into our tax engine, but have not come up with an incentive for them to do so. Certification comes up in this regard, i.e. if you use the state deposited content in our tax engine, there is relief from audit risk. If the states could either deposit content directly (a cost to them) or we could show with high confidence that our analytics engine accurately represented their tax intent, the our platform might be attractive to industry subscribers if audit relief was attached because states had high confidence they were getting their due. This would have to be coupled with a different sort of analytics engine, one that maps the customer’s taxability situation to appropriate rules for the jurisdiction.

  3. Sorry, meant to say above that the value of the tax calculation engine is useLESS without the monthly data update.

Leave a comment