About the Builder
I built Sparse Halo to make advanced AI workflows feel usable, rigorous, and honest.
I'm Ronak Chakraborty. I care about what happens when AI actually has to do useful work, and Sparse Halo is what came out of that. It started as a frustration with single-model tools and turned into a full product with a private chat workspace, a multi-agent debate mode, and an SAT prep tool that runs on the same debate pipeline.
Builder Notes
I'm interested in turning powerful AI systems into tools that people can actually reason with, not just tools that generate impressive first drafts.
At Indiana University's Kelley School of Business, I'm pursuing Supply Chain Management and Business Data Analytics majors with a minor in Public & Environmental Affairs. That mix shapes how I think: in systems, tradeoffs, operations, and real-world consequences. As a college senior, a lot of my experience has come from working on systems that sit between operations, analysis, and execution, whether that meant building a Python-based pre-routing tool at TJX, working in finance and operational roles where clear processes matter, or building independent automation projects ranging from AI infrastructure systems to IPO research workflows. Those experiences are a big part of why I care so much about making technical systems practical, legible, and genuinely useful.
Growing up in the San Francisco Bay Area gave me an early interest in how systems shape everyday life, and I still carry that with me. I love learning, and a lot of my interests live in the space where infrastructure, culture, and decision-making overlap, from public transportation and urban development to vintage watch collecting, music, and personal finance. I'm especially drawn to the kinds of ideas that reward close attention over time, whether that means understanding how a city moves, how a product earns trust, or how a well-designed tool can make complexity feel intuitive.
At a Glance
- Builder
- Ronak Chakraborty
- School
- Indiana University — Kelley School of Business
- Majors
- Supply Chain Management · Business Analytics
- Minor
- Public & Environmental Affairs
Where that perspective comes from
Three disciplines. One design sensibility.
Supply Chain
Studying supply chains made me care about flow, resilience, coordination, and what happens when a system breaks under pressure.
Analytics
Business Analytics sharpened my interest in decisions, structure, and how to make complexity legible without flattening it.
Public Affairs
My minor in Public & Environmental Affairs keeps me thinking beyond the product itself and toward impact, institutions, and practical consequences.
What I built
A dual-mode AI workspace designed and developed entirely on my own.
Sparse Halo gives users two distinct ways to work, depending on what the problem demands.
Chatbot is for direct, fluid conversation with mid-thread model switching. Cabinet is for structured multi-agent work, where one model drafts, another challenges or operationalizes, and the system synthesizes a stronger final answer. SAT Prep is an adaptive practice workspace with AI-generated questions, three AI advisors, and a live score estimate that sharpens with every session.
The goal was never to build another generic chat wrapper. I wanted a workspace that treated different problem shapes differently.
“A quick question, a strategic decision, and a complicated launch plan do not deserve the same workflow.”
The problem
Most AI products lock you into one model and one interaction style. That is convenient, but it also flattens the experience.
I built Sparse Halo around the idea that users should be able to choose the right mode for the task at hand. Sometimes you want a simple exchange. Other times you want pressure, critique, planning, and multiple perspectives stress-testing each other before anything is presented as final.
The Planning persona inside Cabinet came directly from that thinking: not just critiquing an idea, but turning it into a detailed, usable action plan.
The stack
Split-stack architecture, multi-provider model routing.
The browser talks to the frontend through a server-side proxy, which keeps routing and access control centralized. Premium model access is enforced at that boundary and tied to an immutable Microsoft sub identifier instead of something mutable like email.
Next.js 16
React 19 · App Router
FastAPI
Python 3.11 · SSE streaming
Vercel
Frontend deployment
Render
Backend deployment
OpenRouter
Multi-provider model routing
Upstash Redis
Serverless persistence
How I build
Intentional products, honest boundaries.
I care about systems that are honest about their boundaries. That means being careful about what the product really guarantees, where state lives, and which edge cases matter for trust.
Sparse Halo's premium access model is a good example. Authorization is grounded in an immutable identity field because security decisions should rest on something stable, not on a value the user can change.
More broadly, I like building products that feel intentional. The interface, the architecture, and the product promises should all line up.
What's next
Sparse Halo is live at www.sparsehalo.xyz and is still evolving quickly. I am continuing to expand the model suite, refine the collaboration and synthesis pipeline, and explore how structured AI workflows can produce outputs that feel more rigorous than what a single instant response usually delivers.
The long-term goal is to keep pushing on a simple question: how do you make powerful AI systems more useful without making them less understandable?