Like / Dislike **We'd Appreciate Your Opinion**

Leave a Review! http://bit.ly/E6S-iTunes;

Intro: Welcome to the E6S-Methods podcast with Jacob and Aaron, your weekly dose of tips and tricks to achieve excellent performance in your business and career. Join us as we explore deeper into the practical worlds of Lean, Six Sigma, Project Management and Design Thinking. In this episode number 150, we welcome Carlos Fiol to speak about his work dealing with big data text analytics in a default mortgage process. You can find notes and graphics for this episode on website, including a collection of Carlos' wildlife photography. If you like this episode, be sure to click the "like" link in the show notes. It's easy. Just tap our logo, click and you're done. Tap-click-done! Here we go. http://bit.ly/E6S-150 Leave a Review! http://bit.ly/E6S-iTunes

I About Carlos Fiol - fiolc0703@gmail.com

a. Carlos Fiol is a Process Consultant with Wells Fargo. He has over 16 years experience in process improvement in a variety of industries including IT, Manufacturing, Healthcare and Banking with such companies Johnson Controls, General Motors, Christus Healthcare and IBM. Carlos has a BS and an MS in Mechanical Engineering. He is native born in Chile and has a passion for helping others, nature and wildlife photography.

II Very challenging for process improvement professionals to increase performance by only using numerical data

a. Complex Processes in the mortgage industry (servicing loans in default)

i. non-sequential processing (why?)

1. Reinstatement options can be combined in series and in parallel - multiple "endless" process cycles to avoid foreclosure

ii. local and federal government regulations

b. Data Rich, Information Poor-

i. lack of statistical normality

ii. multiple information systems used

iii. 80% of all business information is stored as in free-form text

III Big Data solutions required

a. System events mapping - (step codes) that team members perform on a daily basis

i. Look for common patterns (pareto of process patterns)

ii. Remove a small percentage of the information until clarity is reached

b. Text analytics or text data mining - extracting data from free-form text:

i. (Case notes) Key in analyzing customer interactions

1. events & issues

2. agreements

3. changes

ii. Specialized software products needed (SAS, RapidMiner, etc.)

iii. Develop contrasting patterns

1. group complex information into a single image

2. Pareto issues in processing

iv. Text Graphing Tools

1. Each word-sphere proportional to the frequency / usage of the term

2. Distance between spheres - connection between terms frequently used together

v. Text alignment

1. Identify word(s) of interest and align all sentences by this word

vi. Text Trends

1. Individual loan touch-points

a. qualitative view of customer pain and/or processing fails

IV In combination - Big Date and Text Analytics

a. Cut through the noise

b. Distill common patterns to identify improvement opportunities

c. mine non-numeric data for valuable customer and process insights

V Contact Carlos Fiol: fiolc0703@gmail.com

Outro: Thanks for listening to episode 150 of the E6S-Methods podcast. Stay tuned for episode 151, where we welcome back Johana Ficatier to discuss her insights on career transitions between manufacturing and healthcare. Don't forget to click "like" or "dislike" for this episode in the show notes. Tap-click-done! If you have a question, comment or advice, leave a note in the comments section or contact us directly. Feel free to email me "Aaron", aaron@e6s-methods.com, or on our website, we reply to all messages. If you heard something you like, then Clammr and share it. Don't forget you can find notes and graphics for all shows and more at www.E6S-Methods.com. "Journey Through Success. If you're not climbing up, you're falling down." Leave a Review! http://bit.ly/E6S-iTunes

Like / Dislike **We'd Appreciate Your Opinion**