My advice to young professionals and restarters in IT

My advice to young professionals and restarters in IT. Learn Python both as data scientist tool and web developing tool. Learn JavaScript for a richer web experience. Just don’t go bezerk wanting to learning everything there at once, and, definitely do not put every single technology you encountered in your resume! Don’t always follow instructions. Be badass sometimes – not all the times. Be humble but break status-quo. Be fearless just like you are playing Fortnite. Use Jupyter Labs extensively, enjoy ZSH scripting, try stuff and break stuff. Blog about anything and write something new each day. Have fun even at work. Talk and chat often.  Cry sometimes (we are humans afterall) and cheer at other times. Love what you do and you would do great !

Awesome Interns at Thomson Reuters

Aewesome data science and AI projects with myself and youngest talented member of my team Shiqi (Katherine) Li as intern mentees at Thomson Reuters! – Akhilesh Yeleswarapu and Caleb Fung working on a Javascript & patterns library modules for a tax product, Samir Naqvi working on a interactive chatbot/nlp product for a learning product, Sandilya Madiraju working on fraud detection machine learning tool for an audit product, and Drew Huck working on a cool blockchain identity poc. #TRintern#workingatTR

Sunday notes – blogging on the go

Busy times ahead for me at Thomson Reuters. Before I talk about that, I am just happy that I managed to find a very convenient way to blog post using Jekyll blogging on GitHub while on the go. My lovely daughter allowed me to use her old iPad Mini 2. She got a new Chromebook but she wanted the Mini for watching Netflix while we travel. So we made the arrangement that she uses the Mini whenever she wants. It’s hers after wall but I use it when it is idle… The iPad may be old but it still works like charm. I got myself a Zagg rugged bluetooth keyboard that hooks to the iPad and make the iPad look like a mini computer. I would then be able to carry my “new mini computer” on the go and use it for daily writing, research, and blogging. My big old new Macbook Pro will stay most of the time in my backpack unless I need it to do some really professional coding – something that needs bigger screens and big machines. Other than an old iPad Mini that has now resurrected as a mini computer works so perfectly with what I constantly seek after – a minimalist professional lifestyle.

If you like to do the same, here are the instructions:

  • left over iPad Mini. (Any tablets or phones would) but the size of the Mini is just perfect. Thank you, Apple. Wait! Thank you, daughter !!
  • for note taking on iOS: i tried may apps but I ended up with 1Writer. It is really good and allows you to customize it with Javascript and other techniques. It is my note taking tool of choice for markdown.
  • for blogging: I experimented with different tools and solutions including self-hosting, paid-hosting, etc but I seem to always go back to using Jekyll for microblogging. As for the custom domain, I just purchased my domains from Google Domains and hooked it up to my GitHub repo.
  • for storing files: I wanted an easy way to keep all my notes and files in sync, hence, it was a struggle deciding on DropBox, iCloud, Google Drive, and OneDrive. Believe me when I am saying that I wasted too much time and synchronization effort to decide on what is the best file synchronize strategy for me. Not necessarily for note taking but for everything – this needs a whole blog post. After a millions of trials, I decided to stick to DropBox and live with the 0/month for 2TB that I will only use probably one fourth max. Dropbox teams knows how to price something … no wonder they went IPO. I moved my Zotero-based research papers on DropBox and then updated Zotero settings to allow access to DropBox. I honestly feel bad that I did not go with Zotero own cloud plan because that money goes to developing Zotero, my reference management tool of choice for over a decade. I will donate money to them so hopefully that takes care of my guilt
  • for pushing blog posts from 1Writer app and into GitHub, I found Git2Go iOS Git Client to be a God sent! Thanks to a 1Writer Action for Git2Go I am now able to able to write a blog post like this one, send it to Git2Go app, add the new file, do the commit, and push the post on my blog. Love it

That said, I feel that I am completely mobile now. I can blog. I can write. I can read. I can listen to music. I can research. All with this beautiful device that proves that what Steve Jobs imagined Apple to be has hit home for me.

Dallas ISD Career Fair Student Event

As part of Thomson Reuters community engagement, my team and I met with Dallas ISD high school students who visited the University of Texas in Dallas campus. My team sat with the students and discussed their career experience. I did as well, but I also had the chance to present my deck to the students. Link to the presentation. Hoping to generate excitement to the students about career in computers, I talked about my passion with computer games, highlighted Dallas as one of the founding cities in technology, and then briefly showcased with YouTube videos how AI is being embedded in games. I tried to empower them with the idea that they also can develop cool stuff with AI and software development. After my speech, I sat with some of the students and asked me what they liked about computers. Their answer was what i expected – “games”.

IDEAS 2018 deck: DIY Applied Machine Learning: Practical and Collaborative Method to Jump Start into Machine Learning with Jupyter Notebooks and Google Collab

click here for the deck that I used for the IDEAS – Int’l Data Engineering and Science Association conference in Dallas.

Also here is the abstract that I wrote when I was preparing the material:

Practical and collaborative method to jump start into machine learning projects using open source with Jupyter Notebooks and Google Collab

Not at all machine learning enthusiasts are alike, and, hence, setting up code environment or training the data model can many times be overwhelming for newcomers in the field of machine learning and deep learning. Data scientists may not have the necessary skills in setting up development environments, and, programmers, may not necessarily have the data scientists skills for preparing data sets and evaluating machine learning models. Furthermore, data scientists and programmers together in some enterprise may lack the collaboration platforms to work together on such projects. Even though most of the books on machine learning using some programming language X provide readers with instructions for setting up the coding environment, such chapters can derail the process of getting started if one gets stuck in the setup. Alternatively, collaborative platforms using Jupyter Ipython Notebooks and Google Collab provide a quick starter for programmers and data scientists alike to develop and collaborate on machine learning algorithms through open source without getting stuck with the nuances of setting up their environments or having to depend on commercial products. The talk revisits the value of Jupyter notebooks for newcomers to the field of AI by showcasing live examples and sharing sources of machine learning algorithms running online using Jupyter notebooks and providing a set of guidelines for implementing such technology in the workplace or leveraging existing ones in the market such as Google Collab. The talk is intended to encourage machine learning enthusiasts to enter the field through a practical method that minimizes the stress from the overwhelming material available on the Internet on how to get started with machine learning.

Commodore 64/128 FTP Sites

I found this set of FTP sources for the Commodore 64 and 128 from one old mailing list
Commodore FTP Sites Listing Last update 21 Aug 2009 by Cameron Kaiser


Last updated: 24 Jun 2006
Username: c64
Password: rulez
Directory: /
Description: Various scene and game-oriented files, including an FUnet mirror,
an NTSC demo area, a COMPUTE! archive and a Lt. Kernal archive.

FTP: (
Last update: 01 Sep 1997
Directory: /pub/cpm
Description: CP/M and various system files.
Also available by email from mail-…

FTP: (
Last updated: 16 Jan 2006
Directory: /pub
Description: Demoscene-oriented, a lot of demos, magazines, tools.
Major site for the Plus4. Located in Hungary. Offers C64
and C128 stuff as well. The directory structure is:
Directory: c128 5/31/02 12:00:00 AM
Directory: c64 1/16/06 12:00:00 AM
Directory: 12/31/05 12:55:00 AM
Directory: incoming 4/24/02 12:00:00 AM
Directory: plus4 1/6/06 8:45:00 PM
Directory: users 9/3/05 7:16:00 AM

FTP: (
Last updated: 16 Jan 2006
Directory: /kermit/c
Description: Official distribution site for Kermit, the famous multi-system
terminal program. Offers SwiftLink and standard versions. For
64 and 128. Look for c64*.

FTP: (
Last updated: 16 Jan 2006
Directory: /midi/software/c64
Description: Information on MIDI for the C64 and some software.

FTP: (
Last updated: 6 Dec 2002
Directory: /pub/usenet-by-group/comp.sys.cbm/
Description: Relevant Commodore FAQ documents.

FTP: (
Last updated: 6 Dec 2002
Directory: /pub/c64
Description: Site for c64 software. Very active and currently maintained.
Demo-scene oriented with a large number of demos and coding tools.

FTP: (
Last updated: 15 Jun 1999
Directory: /pub/user/spectre/
Description: The Computer Workshops shareware archive. Various downloadable utilities and games are also in this archive. Most programs are shareware/freeware. Also distributes HyperLink and support files, Kermit for the C64, and various other utilities.

FTP: (
Last updated: 16 Jan 2006
Directory: /pub
Description: This site is mantained by Bo Zimmerman. It contains many Commodore utilities, programs made by a former group, and some programs for BBS’s. Lots of fun little system support utilities, too. And he listens to KISS.
The new directory structure holds the former FUnet content
under /pub/cbm and also some CP/M items under /pub/cpm.

FTP: (
Last updated: 6 Dec 2002
Directory: /pub/c64
Description: Active demo site maintained by Trasher & Riddler of Active. Also lots of emulator-oriented stuff, games, Amiga files and other helpful utilities.

FTP: (
FTP: (
Last updated: 16 Jan 2006
Directory: /pub/c64
Description: Site of the Digital Dungeon (HQ. for SCS*TRC) featuring games, demos, tools, and party pictures. Be sure to
check into all the directories and sub-directories of this
massive collection of CBM files, and the extensive
collection of music files and music tools. Many files are
Unix zipped. Unzip in your ISP shell.

The directory structure includes:
-rw-r–r– 1 1008 1008 1001709 Jan 16 05:07 ALL-FILES.TXT
-rw-r–r– 1 1008 1008 174848 Jan 15 22:46 DEMO122B.D64
drwxrwxr-x 5 1008 1008 4096 Nov 29 07:50 Demos
-rw-r–r– 1 1008 1008 6876 Jan 13 11:36 MISSING_FILES.TXT
drwxrwxr-x 243 1008 1008 8192 Jan 16 11:35 Magazines
drwxrwxr-x 4 1008 1008 4096 Nov 06 13:25 Misc
drwxrwxr-x 10 1008 1008 4096 Dec 15 14:53 Music
-rw-r–r– 1 1008 1008 18501 Jan 16 11:08 NEW-FILES.TXT
drwxrwxr-x 151 1008 1008 4096 Jan 15 00:43 Party
drwxrwxr-x 8 1008 1008 4096 Oct 17 2004 Scene
drwxrwxr-x 20 1008 1008 4096 Aug 24 10:51 Tools
-rw-r–r– 1 1008 1008 1045 Oct 26 14:38 UPLOAD-RULES.TXT
drwxrwxrwt 7 1008 1008 4096 Jan 14 15:13 incoming

IDEAS 2018 Conference Speaker

I will be presenting at IDEAS – Int’l Data Engineering and Science Association Dallas 2018 conference. My topic was just approved by the committee and it is titled “practical and collaborative method to jump start into machine learning projects using open source with Jupyter Notebooks and Google Collab.” The aim of my talk is to encourage more entries to machine learning science using hands-on methods that are more open and collaborative instead of closed black box methods. I know there are more DIY methods and techniques but i am limiting it to few at the talk due to time constraints and respecting audience’s attention. I will however expand the various options in a future paper. You can read more about the Dallas conference at

I hope to see many of you there.

Thank you IDEAS – Int’l Data Engineering and Science Association for accepting my abstract.

Build with Watson

Last night I played around with IBM Watson Internet of Things and Bluemix cloud services platform using their developer tools. IBM changed their developer pricing policy to allow free light accounts not to expire which is a great for unlimited prototyping. I experimented with training Watson conversations dialogues and natural language understanding using their visual aid tool and the github sample codes as part of their machine learning development platform and later ran an IOT simulator using a sample NodeJs app uploaded to their cloud using cloud foundry as part of their IOT development platform. A feature that I liked but didn’t play with it yet was IBM recipes for IoT using NodeRed, a flow based programming tool that was originally created by IBM engineers but is now part of JS Foundation. I plan to a run a real prototype with my Raspberry PIs and Arduinos soon and potentially extend it with Watson as well for an office prototype project.

Check IBM Watson developer program and IBM Watson for IOT

Dallas AI Second Meetup on Deep Learning

In our second meetup event, Viswanath Puttagunta, CTO of Divergence Academy, gave an overview of Convolution Neural Networks, and then Mandeep Kumar, CEO and Mahesh Kashyap CDO of Deep Cognition presented their desktop and cloud-based studio for developing neural networks. Divergence Academy provides courses on various topics including data science. Check Divergence Academy data science program As for Deep Cognition, Inc. they developed a complete platform for designing and applying deep learning algorithms. What I found interesting about their product is that they provide a multi dimensional view of an AI implementation. You first drag and drop building blocks to build a neural network in a wizard-like fashion and then look behind the model to see the generated code. The generated machine learning code is heavily based on ResNets, VGGs, CNNs, and Keras, all very popular deep learning frameworks and libraries. More information about Deep Cognition Thank you Divergence Academy and Deep Cognition for your sponsorship and presentation.

Google AIY

Setting up Google Voice AIY was an easy challenge but getting Google Vision AIY took a deep toll on me to successful solder the smaller Raspberry PI Zero W gpio pins. Messed up two PIs along the way but improving my soldering skills out weigh the cost of two $5 PI Zeros excluding the gas and time lost from the multiple trips to the store. i proudly think that I perfected the art of soldering (personal bias). Nevertheless, the multi faced benefits of Raspberry PIs, along with Arduinos, for the of things multiplied with artificial intelligence are amazing. Couple of years back I won a hackathon thanks to a RasPi idea, now I have one for retro games, and, more recently, couple of ones for AI proof of concept projects. If you never used a Raspberry Pi, go get one ! Any one!

More details about Google AIY is at