Follow the
problem deeper.
I turn ambiguous problems into products, systems, and working software.
Building Ve Browser
Whatever layer it takes.
I don’t sit neatly between the boxes. That’s the point.
I follow the problem through whichever layer it takes to solve it.
User
Listen to the people doing the work.
Workflow
Find the handoffs, decisions and exceptions.
Product
Make the problem tangible and useful.
Interface
Give the system a usable shape.
API
Connect the moving parts.
Data
Make operational context usable.
Infrastructure
Build something that can run.
AI
Explore intent, memory, tools and action.
Working systems.
Real constraints.
Internal tools and product systems shaped around the people who needed them. Four different problems; a shared attention to how work actually happens.
Sales CRM
A working rhythm for leads, conversations, and follow-ups.
Album operations
Connecting the people behind a multi-stage production process.
Asset management
Turning a distributed physical operation into a trackable system.
Workflow & proposal product
From configurable workflows toward a broader media platform.
My career didn’t move from one box to another.
It accumulated layers.
Each new problem added another way of understanding and building. The earlier layers stayed useful: customers, quality, product, operations, software, infrastructure, and now AI.
Questions I’m
working through.
A smaller index of research beyond the professional work. These experiments explore how context, agents and physical constraints change the shape of a system.
Repository harness
Persistent context and engineering standards above individual repositories. The question: how can agents work across projects without repeatedly rebuilding their understanding?
Memory & agent systems
Exploring how intent, remembered context and tools become part of the product itself, including orchestration, MCP and voice-first interaction.
Physical-world systems
Research into livestock wearables, virtual fencing and sensors: useful signals, affordable devices and maintainable systems beyond a screen.
Autonomous trading Experiment ↗
Bounded agent decision-making, evaluation and risk controls. The initial concept was an experiment with approximately ₹1,000.
Terral Exploring ↗
Food processing, sourcing and manufacturing economics: product thinking applied to machinery, supply chains and real-world operations.
The roles changed.
The pattern didn’t.
I started close to customers and support. Each time I encountered a problem, I moved one layer deeper — testing, product, internal systems, development, infrastructure, and now AI.
I kept following the problem.
I like turning ideas into systems.- Understand the workflow.
- Build before over-planning.
- Stay close to reality.
UnderstandFind the real problem.↗
Start close to users and the work outside the interface.
ModelMap it. Simplify it.↗
Follow the roles and constraints. Find the smallest useful intervention.
BuildMake the idea real.↗
Prototype and implement the system so its assumptions can be tested.
ShipPut it into use.↗
Deployment and operations are part of making a product useful.
ObserveLearn. Return. Iterate.↗
Watch what happens in practice. Use feedback to improve the system.
Let’s build something
interesting.
Products. AI. Systems.
Strange problems worth solving.