#286: Getting Going with Machine Learning for Finance and Accounting with Ashok Manthena

Our Guest Mentor:                                                        

Ashok Manthena is an experienced Product Manager and Systems Integrator with a history of working in the information technology, Retail, Manufacturing industry. Expert in Data Analysis, Business Analysis, Financial Systems with strong financial professional background which he earned from his stints at Gap, Google and Ingersoll Rand.

Ashok Manthena is based out of San Francisco, California, USA and has an MBA from The University of Akron.

 Key Quotes From the Episode:

“Forecasting in general is a guess work. It’s not going to be 100% accurate.”[08:28]

“It’s about selecting the right model that applies to your data then fine tune it to give you the best output.”[17:10]

“We always judge ourselves by intent and we judge others by their action.”[28:08]

 (Other resources and how to connect with Ashok below).

Get Show Notes and Links in your Inbox!


Key Points From the Episode:

– Why Start-up companies adapt to machine learning technology faster and why mature ones are reluctant to invest in it

– How machine learning makes forecasting easy and why an appropriate rate of accuracy is an important consideration

– And different machine learning models to get a start on in Finance

o Simple regression



Time Stamped Show Notes

[03:18] – Ashok shares his career journey working in financial system and ERP. Providing services to many financial and accounting clients which before many where doing the manual processes but shifted to automation lately.

[04:06] – Start-up companies adapt to technology faster compare to already established companies.

[04:54] – Established companies are reluctant to invest in machine learning technology

[06:05] – Four years of stored data is important for machine learning

[07:41] – The future of machine learning for finance and accounting looks bright

[08:15] – Machine learning makes forecasting easy

[09:41] – Accuracy of forecasting using the technology

[11:50] – Above 95% accuracy of forecasting is good or any figure you are comfortable with

[11:42] – Accuracy rate for current system used on their forecasting activities

[12:22] – Best place to connect with Ashok and information about how machine learning works

[13:25] – Simple ways to start using machine learning, time series forecasting

[15:19] – Forecasting not every single chart of accounts

[16:18] – Different machine learning models;

o Simple regression



[18:18] – Think about the machine learning part as experiential for us to know what works and what doesn’t

[19:41] – Hitting your sweet spot after constant use of machine learning

[21:12] – Building your own machine learning vs. using available software online or from vendors

[23:20] – Customization of machine learning tool to match the functionality to the company’s needs

[25:21] – Ashok’s shares his favourite application from machine learning successes

[27:27] – Ashok best advice ever received.

[28:41] – Breaking silos in the organization including the technology it adopts

[29:43] – Ashok parting thoughts on optimizing ourselves

Connect with today’s guest:

LinkedIn : https://www.linkedin.com/in/ashokman/

SUBSCRIBE ON iTunes & Android