https://shashankkumar.pages.dev/Recent content in Shashank Kumar Shankar on Shashank Kumar Shankar | Product Manager, Cloud & AIHugo -- gohugo.ioen-usSat, 19 Sep 2026 00:00:00 -0700https://shashankkumar.pages.dev/posts/jev/Sat, 19 Sep 2026 00:00:00 -0700https://shashankkumar.pages.dev/posts/jev/Which team should handle this ticket? Which tool should an agent run next? Is there enough information to proceed? A lot of AI application work comes down to small decisions like these. What caught my attention about Jev from TypeSafe AI is that it makes those decisions the product: give it context and defined questions, and get structured answers that code can act on. After Hebbia, I wanted to look at another company changing how software uses AI.https://shashankkumar.pages.dev/projects/Fri, 18 Sep 2026 00:00:00 +0000https://shashankkumar.pages.dev/projects/I build AI apps, interactive explainers and browser games alongside my work in cloud product management. These are personal projects I design and develop with AI coding assistants. Apps · Interactive explainers · Games and visual experiments · Earlier projects Apps AI Trio Put a question to three open AI models in one conversation. They can respond to each other, challenge assumptions and surface different ways to approach a problem.https://shashankkumar.pages.dev/posts/building-ai-trio/Tue, 03 Feb 2026 06:10:01 +0000https://shashankkumar.pages.dev/posts/building-ai-trio/I kept running into the same frustration: every LLM has blind spots. Ask GPT a question, you get GPT’s particular style and biases. Same with Claude, same with Gemini. The answers aren’t wrong, but they’re shaped by each model’s training, tendencies, and assumptions. I wanted a tool where I could ask all three the same question simultaneously and watch them debate each other - like a group chat where AI models challenge each other’s reasoning.https://shashankkumar.pages.dev/posts/product-insights-1-hebbia/Thu, 20 Nov 2025 06:10:01 +0000https://shashankkumar.pages.dev/posts/product-insights-1-hebbia/I am starting a series where I break down AI products to understand what they are building, why it matters, and whether the business makes sense. First up: Hebbia - The AI platform for finance. The problem: An auditability crisis Here’s what makes Hebbia interesting. They are not solving a “search” problem. They are solving what I’d call an auditability crisis. Private equity firms and law firms can’t just throw documents into ChatGPT and trust whatever comes out.https://shashankkumar.pages.dev/posts/thoughts-on-product-management/Mon, 12 Feb 2024 06:10:01 +0000https://shashankkumar.pages.dev/posts/thoughts-on-product-management/My previous blog post was almost 2 years ago on my thoughts on why I am switching my career from engineering to product management. Now, its been 2 years into the PM job while surviving layoffs in the tech industry (till now atleast). Given that, I have some perspectives on this role (mostly at Microsoft) and what it could be going forward. Product management experience so far Product management is very different across different companies.https://shashankkumar.pages.dev/posts/switching-to-product-management/Mon, 18 Apr 2022 06:10:01 +0000https://shashankkumar.pages.dev/posts/switching-to-product-management/Since Jan 2022, I switched careers from being an SDE to PM. I wanted to share my thoughts here on “Why” did want to switch careers: What do I want to do in my career? As of July 11th 2022, I would hit 6+ years in Software Engineering career. I would have learnt a lot in semiconductors (Intel), SaaS (Workday), ML platforms (AWS). I would be at a point in my career where I have been good at how to build a product.https://shashankkumar.pages.dev/photography/Tue, 13 Oct 2020 07:58:41 +0000https://shashankkumar.pages.dev/photography/ I capture the intersection of Abstract, Nature, and Urban life through my lens. Open full portfolio in new window →https://shashankkumar.pages.dev/posts/started-tech-startup-venture-podcast/Sun, 04 Oct 2020 06:10:01 +0000https://shashankkumar.pages.dev/posts/started-tech-startup-venture-podcast/[Update as of April 2022] Shutdown the podcast because of following reasons: Super tedious to edit / create videos + audio. Less viewership. Didn’t enjoy doing it anymore because of 1 and 2 where the feedback loop to do it further was not there. Learnings from running the podcast and shutting it: Creating media content which users want to listen / view is pretty hard.https://shashankkumar.pages.dev/posts/naval-ravikant-how-to-get-rich/Tue, 18 Feb 2020 06:10:01 +0000https://shashankkumar.pages.dev/posts/naval-ravikant-how-to-get-rich/Amazing series of tweets on wealth generation by Naval Ravikant who is the founder of AngelList and an investor in a host of successful companies. The podcast discussing these tweets: And the link to the original thread of tweets: Seek wealth, not money or status. Wealth is having assets that earn while you sleep. Money is how we transfer time and wealth. Status is your place in the social hierarchy.https://shashankkumar.pages.dev/posts/overview-ml-crash-course-1/Mon, 12 Aug 2019 06:02:46 +0000https://shashankkumar.pages.dev/posts/overview-ml-crash-course-1/Helps do 3 things better for Software Engineers: Reduce time programming: Move from rule based program (logic) to writing programs where we feed it a data and we get results from it which are closer to the result. Customize products for specific users: Ex: Instead of writing two programs for English / French users, just write a general ML program and feed it English / French to generate language specific results.https://shashankkumar.pages.dev/posts/overview-learning-how-to-learn-coursera/Fri, 05 Apr 2019 06:59:15 +0000https://shashankkumar.pages.dev/posts/overview-learning-how-to-learn-coursera/This is my notes from the course Learning How to Learn from Coursera Quick look at concepts consolidated from the course Focused mode vs diffused mode 25 minute Pomodoro method Long term vs Short term memory Spaced repeated practice to move from Short to Long Sleep well Chunking Releated practice Mini-testing Recall Illusion of Competence Overlearning Deliberate practise Einstellung Interleaving Procrastination Procrastination as a Habit Habit made of cue, routine, reward, belief Initial start is hard Process vs Product Use auto-drive zombie mode (blindy do) Activity tracking before sleep for next day Pomodoro early in the morning Conjure visual images to remember General Problem solving: Apply known technique Modify known technique Focused brute force thinking Diffused related thinking Repeat until solution Exercise best method for brain enhancement Visual metaphor (Mud + JSON Marshall) More Genius less creative Imposter Syndrome perseverance = the virtue of the less brilliant Teamwork for fresh brain perspective Minitest benefit »> Learning benefit Left brain vs right brain (focused vs diffused) Checklist to measure test / interview preparation quality Start hard, jump to easy (Focused vs diffused) Anxiety = Excited for test, not afraid Breath slowly Week 1: What is Learning?https://shashankkumar.pages.dev/posts/swe-interview-and-learning-tips/Mon, 01 Apr 2019 07:25:15 +0000https://shashankkumar.pages.dev/posts/swe-interview-and-learning-tips/I was watching through Joma’s video on his YouTube channel about a guy named Cyrus Xi receiving amazing offers and I think his tips on preparing / learning in general are worth documenting. So, here it is. Video link: His reddit posts: Pre-FB: https://www.reddit.com/r/cscareerquestions/comments/b3w1gg/2_year_update_my_journey_tips_29_gpa_at_a_small/ Currently after getting Uber, Snap etc.: https://www.reddit.com/r/cscareerquestions/comments/9i0pow/graduated_with_a_24_gpa_1_year_ago_after_studying/ His study guide: https://workflowy.com/s/study-guide/RD5kZ682pWX5oxiE Along the read, I found this amazing article on Career Planning. Seems it was originally written by Marc Andreessen.https://shashankkumar.pages.dev/posts/software-product-management-uc-berkeley-ext/Wed, 27 Mar 2019 06:07:19 +0000https://shashankkumar.pages.dev/posts/software-product-management-uc-berkeley-ext/I took this course as a part of Spring 2019 semester coursework at UC Berkeley extension. I see soo many engineers jumping on the Product bandwagon. I signed up for this to get a sense of what the fuss is all about. UC Berkeley extension is a program from UC Berkeley where individual courses from various departments can be taken as evening sessions in SF or Berkeley campus. Seems like it is targeted for folks looking to learn a new skill or for a career switch.https://shashankkumar.pages.dev/resume/Sun, 17 Feb 2019 07:58:41 +0000https://shashankkumar.pages.dev/resume/Bellevue, Washington · shank4804@gmail.com · linkedin.com/in/shank7485 Download PDF Senior Product Manager with 9+ years spanning product management and software engineering at Microsoft Azure, AWS, Workday and Intel. Led the 0-to-1 build and commercialization of Azure Storage Actions - now automating data management for 300+ global enterprises handling 2B+ objects. I pair deep familiarity with distributed systems with hands-on GenAI and LLM agent prototyping to ship platform capabilities, craft monetization models, and guide cross-functional teams from concept through global launch.https://shashankkumar.pages.dev/contact/Sat, 16 Feb 2019 08:24:42 +0000https://shashankkumar.pages.dev/contact/ Email: shank4804@gmail.com LinkedIn: linkedin.com/in/shank7485 Twitter: twitter.com/sshank_ GitHub: github.com/shank7485https://shashankkumar.pages.dev/posts/unit-testing-in-golang/Fri, 16 Feb 2018 08:38:00 +0000https://shashankkumar.pages.dev/posts/unit-testing-in-golang/I recently started writing some code in Go (or Golang?) there are lot of amazing things to explore. Its always exciting to learn a new language. So far, I absolutely love Golang. If you are like me who likes C, Python and all this systems/low level OS stuff, you’ll get a feeling that Golang for systems will be similar to what JS is for the web. (+ve JS things :-) ).https://shashankkumar.pages.dev/posts/application-on-kubernetes/Tue, 12 Dec 2017 07:49:08 +0000https://shashankkumar.pages.dev/posts/application-on-kubernetes/It’s a world of microservices. With soo many companies containerizing their apps and scaling them using Kubernetes, it’s hard not to ignore. To an idea of Kubernetes and to play around with it, I created this tutorial to install and deploy dockerized apps on a kubernetes cluster. This tutorial has 3 parts: Dockerizing an API Deploying a Kubernetes cluster Deploying the Dockerized API on the cluster and exposing it.https://shashankkumar.pages.dev/posts/vim-auto-complete/Wed, 10 Feb 2016 07:48:15 +0000https://shashankkumar.pages.dev/posts/vim-auto-complete/Get code completion for Python like an IDE in the terminal using Vim editor with YouCompleteMe and Vungle plugin. Similar to This is for Ubuntu Linux (Debian) based OSes. Steps to be followed: Update to latest Vim. This is done as Vim version greater than 7.41 isn’t available in Apt-get of Ubuntu Linux. sudo add-apt-repository ppa:pkg-vim/vim-daily sudo apt-get update sudo apt-get install vim dpkg -s vim | grep 'Version' #should be greater than 7.https://shashankkumar.pages.dev/posts/irc/Fri, 08 Jan 2016 07:46:15 +0000https://shashankkumar.pages.dev/posts/irc/Ever since I wanted to contribute to an open-source project, I was told on online forums to contact the specific IRC channels to ask questions and start contributing. That’s fine. I’ll do that. But I never understood what was IRC. So what’s IRC? Wikipedia quotes ‘Internet Relay Chat Protocol (IRCP) is an application layer protocol that facilitates communication in the form of text. The chat process works on a client/server networking model.https://shashankkumar.pages.dev/about/Mon, 01 Jan 0001 00:00:00 +0000https://shashankkumar.pages.dev/about/I’m Shashank, a Senior Product Manager on Azure Blob & Data Lake Storage, based in Bellevue, Washington. At Microsoft, I led Azure Storage Actions from its early concept through commercialization and global launch. The work spanned customer discovery, product requirements, pricing and metering, and helping customers put the platform to use. It now serves 300+ enterprises managing more than two billion objects. From engineering to product I started as a cloud software engineer at Intel, then built distributed systems at Workday and machine-learning infrastructure at AWS SageMaker Ground Truth.