Indian quantum computing company QpiAI has launched a finance-focused software development kit (SDK), along with a post-quantum cryptography (PQC) product and a quantum random number generation (QRNG) service, at the Singapore FinTech Festival (SFF) 2026.
The QpiAI Finance SDK is a Python library for modelling, solving and evaluating optimization and simulation problems using classical, quantum-inspired and quantum methods, according to the company. QpiAI says it is aimed at banks, asset managers and trading desks and can run on conventional infrastructure, with the option to move workloads to quantum processors such as its 64-qubit Kaveri QPU.
Quantum-inspired optimization
QpiAI says exact classical solvers struggle with problems involving thousands of discrete decisions under risk, liquidity and transaction-cost constraints. The SDK compiles financial models into Quadratic Unconstrained Binary Optimization (QUBO) or Higher-Order Binary Optimization (HUBO) formulations and solves them with quantum-inspired methods on classical hardware.
The company cited benchmark results for its underlying QpiAI-Opt engine against a classical baseline. These figures were supplied by QpiAI and have not been independently verified:
- Portfolio construction: a 400x speed-up in the discrete optimization step for an asset selection case with more than 15,000 variables.
- Hedge and rebalance selection: a 21x lower time-to-solution and a 1% better objective on a 500-variable instance.
- Payment rail routing: a 5.4% better objective in 5.2 seconds on a 5,000-variable problem covering 1,250 transactions.
- Constraint verification: 100% compliance with rules such as one choice per entity, using the library’s built-in evaluation workflow.
Modules and algorithms
The SDK includes ready-made models for portfolio construction, capital management and risk diversification, with constraints on leverage, cash, holding period, turnover and discrete trading lots. QpiAI also lists support for credit funding allocation, asset-liability matching, hedging, payment settlement scheduling and collateral optimization.
Users can solve these formulations with classical linear, quadratic and mixed-integer methods, or with gate-based quantum algorithms including QAOA and VQE and their variants. Quantum execution is supported through QpiAI-QCloud simulators and the company’s quantum processors.
HERMES framework for risk estimation
QpiAI also described QpiAI-HERMES, a quantum algorithm framework for financial risk estimation, scenario-weighted valuations and stress testing. The company says it moves stress-weighting and linear-combination-of-unitaries (LCU) expectation value calculations to classical computers, leaving the quantum processor to run short measurements. QpiAI claims this requires no extra qubits or added two-qubit gates and produces circuits up to 99% shorter in benchmark tests.
QRNG and post-quantum security
QpiAI has released a fully managed QRNG service that it says integrates with digital workflows and includes verification protocols so enterprises can audit remotely generated quantum randomness. The company positions it for uses such as Monte Carlo pricing and cryptography, where randomness quality matters.
Separately, the QpiAI PQC Transition Suite is intended to help institutions assess, secure and deploy quantum-resistant infrastructure without full replacement. QpiAI says it supports hybrid migrations and covers network communications (TLS, VPN, IPsec), data center storage and authentication, using NIST-standardized algorithms. The company framed the product as a response to “harvest now, decrypt later” attacks, in which encrypted data is collected today for decryption once sufficiently powerful quantum computers exist.



