nihar shah
homeblogresearchcontactResumeGitHub

Research

I tend to research problems I encounter while building.

Google Scholar

BALI: Solvency-Constrained Batch Liquidation as a Cryptoeconomic Control Mechanism

How should liquidation systems behave when solvency matters more than individual positions?

SECRYPT 2026 (full paper) · 2026

Calibrated Agents as Market Makers: Automated Liquidity Provision for Long-Tail Prediction Markets

Can an agent hold a market open where no human would bother to quote?

IEEE AI-R2D2 2026 at ICBC (full paper) · 2026

Hyperbolic Funding Rate Mechanisms for Open Interest Balancing in Perpetual Futures Markets

Can market incentives rebalance risk before intervention becomes necessary?

Designing DeFi 2026 · 2026

Information-Theoretic Limits of Settlement Verification in Oracle-Based Decentralized Finance

How much can verification actually establish before something has to be trusted?

BSCT 2026 · 2026

Settlement Verification in DeFi Credit: An Empirical Comparison of Production Oracle Systems

What does it actually mean to trust an oracle when money settles against it?

IEEE CryptoEx 2026 (poster) · 2026

Bootstrapping Risk-Bearing Capital: Solver Economics for Coordinated Liquidation in Onchain Perpetual Markets

Who takes the other side of risk when liquidation has to be coordinated?

Preprint (SSRN) · 2026

Tail-Risk Control for Side-Effecting Agents Under Scarce Oversight

How do you know an autonomous system did what it intended to do?

Preprint (SSRN) · 2026

get in touch

githubscholarxlinkedinemail

If you're working on something hard to explain in one sentence, I'd probably like to hear about it.