Preparing for leaving Amdocs post

content/posts/2024-end-of-an-era-thank-you-Amdocs/index.md

@@ -0,0 +1,56 @@
+---
+title: End Of An Era - Thank You Amdocs for Amazing 30 Years!
+author: Adi Rabinovich
+type: post
+heroStyle: "background"
+showTableOfContents: true
+draft: true
+date: 2024-07-31T11:11:11+00:00
+url: /2024/thank_you_amdocs_for_amazing_30_years/
+tags:
+  - Career
+  - LinkedIn
+  - Civilization
+  - Open Source
+  - Programming
+  - AI
+  - Amdocs
+  - Vubiquity
+  - Netflix
+
+---
+
+## Thank You
+
+WOW - I am truly at loss for words. Most posts just pour out of me, but this is the toughest one yet. 
+Over my 30 years in Amdocs I was blessed to meet so many talented people. I am especially grateful that they were able to embrace me for who I am. Throughout my career I saw again and again how people from all over the world come together and create amazing things. And especially in the last 10 month - when there are forces focused on sparking division, when my identity is challenged, it is deeply reassuring to see people of different backgrounds world-wide continue to love and support each other.
+So many feels (as youngsers say) - but most surprisingly, Guilt! I realize I am being selfish in this departure from Amdocs. My love for the company runs deep, but I am comforted that there are many talented people who will carry on the torch and will deliver on all our promises!
+I will miss everyone endlessly, and I know our paths will cross again as we join forces to uplift humanity through technological revolution.
+
+## Ab Initio
+
+I was fortunate to join Amdocs while still studying at Israel's Open University. However at the time already had experience working in a computer sales/service firm for several years, and later in a small 3 people startup trying to create an efficient Medical Records system for Israel and beyond.
+Soon a call came in for relocation to USA to support Amdocs client on-site. It was the time of Borland C++ connected to larger servers (VMS !). One interesting challenge I remember back then was to embed Web Server screens (and APIs) into otherwise native Windows GUI C++ application. I got to tinker with HTTP protocol at TCP/IP level, to ensure APIs are dynamically activated properly in various scenarios. Another fun challenge was to develop a self-contained DLL with a lot of business logic, for mobile sales force. I started with writing reference-counting garbage collector which really made the rest of the project easy and efficient afterwards.
+In 2001 Amdocs offered me an opportunity with BellSouth in Atlanta. Probably the biggest highlight there was with small 3-people innovation team that I lead. We reimagined Amdocs older Sales UI as a high-speed Adobe Flex (now Apache Flex) App, with embedded DB for offline support. UI was fluent and intuitive, which sales reps loved and were even able to show-off to their clients during the sales process.
+
+## 3G, 4G, 5G and Beyond - AT&T Mobility
+
+In 2009 I heard of an interesting OSS project starting at AT&T mobility (via Amdocs). I was fortunate to join and even though environment was very competitive and high-stress, the many challenges with a mix of new technologies made up for it. By that time I was already a big fan of Java, and with that opportunity we were able to push the tools to their limits. The endless challenges of ever evolving network technologies, coupled with the high-volume throughput and stability demands, really kept things interesting for the next 11 years! Probably highlight there was designing and implementing a microservice with built-in DSL to support delcarative network design (with some YANG similarity, but extending beyond it).
+
+## Event Horizon
+
+<img style="float: right; margin: 0px 0px 10px 10px; display: inline;" src="man_at_career_crossroads_dalle3.png" alt="Man at Career Crossroads Pixelated" width="200" height="200" align="right"/>When people ask me how can one remain at the same company for 30 years - I always point out that Amdocs is just an Ocean of opportunities. At any time there are clients seeking additional help, so if one gets bored at specific assignment, internal move is easy enough to orchestrate.
+In 2022 I decided to try my hand in something new by joining Vubiquity, one of Amdocs aquired companies. While they do many things to support Entertainment industry, my focus was around cloud-based white-label Entertainment Suite solution. Think Netflix-like solution, including Apps on various platforms, all seamlessly connecting consumers with their favorite streaming videos from Hollywood (and smaller) studios. While undoubtedly video streaming has came a long way by, it is amazing how many technological challenges remain in smooth video delivery, especially for international audiences with many languages and Ads.
+As I celebrated 30 years with Amdocs, thoughts started to creep in to consider exploring what else is out there. I call this - Midlife crisis of sorts. I especially wanted to see if I can contribute to society in a more meaningful way, and perhaps give back to Open-Source community. Throughout the years I wished to give back on several occasions, but often complex licensing agreements and NDAs made it prohibitive.
+I also eternally fascinated by the rapid progress in AI in recent years. I worked with neural-networks over 30 years ago in Tel-Aviv University, but felt there wasn't enough time for deeper dive into latest advancements with full-time work. My exposure while solutioning was limited, mostly as prompt engineering or integrating to existing LLMs and other models. I am fascinated and hopefully to now have the time to dive much deeper. We truly just scratching the surface with AI, especially when coupled with robotics, it will benefit humanity in increddible ways.
+
+## Some related links to explore
+
+- [Event Horizon](https://en.wikipedia.org/wiki/Event_horizon)
+- [Bing CoPilot AI](https://bing.com/chat)
+
+{{< alert "image" >}}
+**Images By DALL-E 3 from Microsoft Bing**
+{{< /alert >}}
+
+![Man looking ahead to continue technology revolution](silicon_career_crossroads_starships_dalle3.png "Exciting Technological Advancements Ahead")

content/posts/2024-swimming-in-dark-pools/index.md

@@ -21,20 +21,15 @@ I have been thinking about algorithmic trading for a number of years. I enjoy ex
 ## Introduction to Algorithmic Trading and Quants 
 
 
-our household this is a common occurrence as well - consoling and supporting each other as we hear the devastating news of yet another Netflix show cancellation.
-Most recently my 10 year old had this experience for the first time also, as I had to let him know that his favorite Dead End: Paranormal Park got axed. There were a lot of tears and screams of "Why!?", so I set out to explore this further. Specifically, how can it be that Netflix, which is (arguably) at the forefront of Big Data Analytics and "Data Driven Decisions", can get this so wrong?
+## Dark Pools - What are those?!
+
 
 ## The young "Flash Crash Trader"
 
 
 https://www.bbc.com/news/explainers-51265169
 
-Statistical analysis is very easy to skew into "expected results". This is known as Confirmation Bias - often Data Analysts will unconciously (or even conciously) slice and aggregate their data to fit their theories. With Watching data I can see where it is infinetely harder, because people watch in so many different ways and patterns. Some bindge entire seasons, others watch episode or even part of an episode at a time - in sessions that may have days/weeks between them. Is it because show is uninteresting or because they have busy lives?
-Another huge factor is Exposure - Can't watch something you never heard about! When Netflix started producing their own content - their biggest issue became making consumers Aware that show/movie exists! They didn't yet have any venue for "coming soon", there were no Ads in their own content, to cross-promote. Really they had Nothing - yet they somehow expected consumers to all Jump on new Original Content and when they didn't - conclusion "Bad Content!". I loved Sense8 and The OA, but by the time I discovered that they exist - they were Long Cancelled!
-A very large factor in distorted results is incomplete or inaccurate data collection. Are they correctly collecting viewership on all devices - what if it is casting from phone to SmartTV with some exotic OS? or Nest device with a screen? And what about the all too common situation where content keeps playing long after the viewer fell asleep?
-A related phenomena I saw first hand is in Ad business - most Ads, even those clicked on, are "false positives". Advertisers (on DSP) know that "exposure" data is flawed - if you start asking consumers who supposedly saw your Ads what do they remember - majority don't remember anything, not the product name nor the company name...
 
-## Leaving Soon - Watch Now!
 
 Another interesting debate are movies/shows that are ending their licensing agreement - this means Netflix losing access to it. Somehow Netflix thought it is best to hide such important information. This is a bit like dirty laundry - you don't want your customers to be aware that they are about to lose access. On the other hand, they may be even more distraught to discover movie disappeared from their watchlist (or stayed on the list, but unwatchable - HULU?!?). Of course executives hope that people don't even notice that their movie is now gone, but in reality many end up upset at unexpected loss.
 I really wish they instead embrace the beautiful "Leaving soon, Watch Now!" warning - so favored by Tubi nowadays (to the point where shows leave just for the ratings boost, and then a bit later come back on the platform...)