Everyday Reflections

The year, as I think, has involved increased efforts towards my running goals. It also included an insane amount of professional travels. I did spend quiet moments with the family whilst as well exciting ones as a family on our Scandinavian tour.

My Key takeaways from 2024…

Through coaching and reading , I have found a lot that has become more relevant to me at this stage of my life after spending almost 23 years in building products and platform services

Active Listening Changes Team Dynamics
Life is all about learning and growing. By truly hearing my team, we recognize not just their words but their unspoken potential. This understanding helps us create more adaptive, resilient engineering cultures

Routines Aren’t Restrictive. They’re Empowering.
Routines are important and getting back to them is even more important!!!” This applies equally to engineering workflows and personal development. Keep a casual tab on it. Do not get paranoid about it. Build memory muscle to power routines. Avoid using pedantic micromanagement to drive them.

Iteration Is Our Superpower
Most profoundly, I realized that “Keep iterating. Stay persistent. Together, they form the cornerstone of impactful problem-solving!” This is important across personal or professional life. Take it in small chunks. Create a habit around breaking down your outcomes. This way, you can make way through your aspirations. Again without persistent breaking down will not help as you end up with glass half full!

Fitness and 2024…

It has been a good year for me from a fitness point of view , covering 1200 kms. I participated in around 15 half-marathons. My top run was 25 km. I look back at a satisfying year of applying my key takeaways to myself!

My take on AI…

How can I not have one 🙂

It has been a crazy year of generative AI. Deep research and exploration are now affecting various facets of our personal and work life.

It brings productivity in certain areas of your work. Being suggestive is what I find to be a key differentiator. Whilst agentic flow become more mainstream, engineering will shift from creating a solution to developing an AI agent. This change signifies the real shift for software engineers on what they are building in future than fearing their replacement.

Above all context is important and owning what you query is important. The query interfaces that drive your context become essential. They allow leveraging generative as core-infra. This is not straight out of the box. We need to codify our thoughts. Each of us is unique, and this uniqueness makes us different after all!

Things to look out in this space would be

NotebookLM : It is a great research and note taking tool. I have been using it off late to do a lot of work. It helps me understand how to gain productivity when dealing with multiple sources of information. The podcast generating feature is great. Recently, your ability to interject a podcast is even greater. You can help add yourself to the podcast interaction!

MCP – Model Context Protocol(MCP) : Anthropic has open-sourced the Model Context Protocol (MCP). It serves as a standard for connecting AI assistants to various data sources. MCP aims to improve AI responses by eliminating the need for custom integrations with each data source. It offers a universal protocol, allowing developers to build secure connections between AI tools and their data. The protocol includes specifications, SDKs, and pre-built servers for popular platforms. Early adopters are already using MCP, highlighting its potential to enhance AI functionality and streamline development. The project is open-source, encouraging community contributions to build a more context-aware AI ecosystem.

PydanticAI – A Python Agent Framework for Generative AI : PydanticAI is a new Python agent framework. It is built by the Pydantic team. The framework is designed for creating production-ready applications using generative AI. It offers model-agnostic support for various LLMs. It integrates seamlessly with Pydantic Logfire for monitoring. Additionally, it features a type-safe design using Pydantic for structured responses and data validation. The framework employs dependency injection for easier testing and features streamlined response handling.

& finally something that I took as a viral take from LinkedIn was coauthor.studio doing a summary of your LinkedIn in 2024 🙂

Looking ahead to 2025…

I’m committed to scaling not just technological capabilities, but human potential. For every engineer, leader, and innovator feeling uncertain – your next breakthrough is just one persistent iteration away. On fitness side I will try to mix my activities more , than just having a personal bias for running! Let us see how things pan out in 2025 !

Wishing each one of my readers a very happy new year ! Stay fit , Healthy and Happy !