Billing Intelligence Platform
Unified fragmented finance events into a single operational view for product and revenue teams.
Next.js / TypeScript / Postgres
2026
Chirag Poornamath · Portfolio
Building systems
with durable intelligence.
I architect product systems where AI orchestration, distributed reliability, and interface clarity are designed as one cohesive operating layer.
Currently building
Agentic workflows for high-context enterprise operations.
Featured Work
ZeroTrace
Privacy-preserving quadratic funding platform
A decentralized platform enabling fair funding of public goods using quadratic voting, zero-knowledge proofs, and sybil resistance.
- Anonymous voting using zero-knowledge proofs
- Quadratic funding prevents whale dominance
- Sybil resistance via Anon Aadhaar
Solidity, Next.js, ZK Proofs, IPFS
Live preview unavailable? Open in new tab
Project Index
Onboarding Flow Redesign
Reduced time-to-value with a guided setup experience and adaptive progression model.
React / Tailwind / Framer Motion
2025
Developer Analytics Console
Built a high-density observability surface with drill-down paths for engineering leadership.
Vite / Node / ClickHouse
2025
Enterprise Permissions Model
Designed role architecture and policy controls to support multi-team account structures.
TypeScript / GraphQL / Redis
2024
Case Study
Problem
TARS AI needed to orchestrate high-context requests across multiple enterprise workflows without degrading response quality, latency, or policy compliance. The existing setup relied on disconnected prompts and static logic, producing brittle outcomes at scale.
Approach
We reframed the product as an execution system rather than a chat surface. Requests were decomposed into intent, context retrieval, planning, and action phases with explicit contracts between each stage. This created a deterministic backbone while preserving language-model flexibility at decision points.
Architecture
The platform combines a central orchestration layer, persistent memory abstraction, and a task runtime that dispatches tools with guardrails. State transitions are observable and replayable, allowing teams to inspect every intermediate decision and tighten reliability through iterative policy tuning.
System Flow
- User
- LLM
- Memory
- Task Engine
- Output
About
I design and build engineering systems where product clarity, distributed architecture, and AI capabilities reinforce each other. My focus is shipping platforms that stay reliable under scale while remaining simple for teams to operate.
Across system design, agent orchestration, and backend workflows, I care about strong interfaces, measurable performance, and long-term maintainability. The goal is practical intelligence: software that reasons well, coordinates well, and delivers value in production.
Contact
If you're building systems that demand intelligence, scale, and reliability — I'd like to be part of it.
Based in Mumbai · Open to remote
Resume