Engineering @ TCS | Prev @ Deloitte | IBM Quantum Qiskit Advocate | Independent Researcher

github

medium

projects

Independent Researcher & Software Engineer

Soumyadip Sarkar • Peerlist Profile

Clustor is a high-performance clustering toolkit that leverages the power of Rust to deliver fast and efficient clustering algorithms. Designed with a Python-first API, Clustor integrates seamlessly into production data pipelines, offering minimal-dependency implementations of common clustering algorithms. Built on PyO3 and maturin, it supports a range of unsupervised learning primitives, making it an ideal choice for data scientists and developers working with large datasets.<h3>Highlights</h3><ul><li>KMeans and MiniBatchKMeans with KMeans++ initialization, Euclidean/Cosine metrics, and streaming partial_fit.</li><li>BisectingKMeans for hierarchical clustering.</li><li>DBSCAN with noise labeling and core sample indices.</li><li>OPTICS for reachability ordering and core distances.</li><li>Affinity Propagation for exemplar-based clustering.</li><li>BIRCH for streaming clustering using CF-tree summaries.</li><li>GaussianMixture with model selection metrics like AIC and BIC.</li><li>Cluster validation metrics: Silhouette, Calinski–Harabasz, and Davies–Bouldin.</li></ul>Clustor is actively developed, with APIs subject to change before the 1.0 release. It is open source, licensed under the Apache License, and available on <a target="_blank" rel="noopener noreferrer" class="break-all link-break text-gray-500 hover:underline" href="https://github.com/alphavelocity/clustor">GitHub</a>.

clustor

Qiskit Inspect offers production-ready debugging, tracing, and analysis tools for Qiskit, enhancing the capabilities of Qiskit v2.0+ with practical features. This tool is designed to assist practitioners in debugging classical control flow, validating sampler output against exact statevectors, and ensuring robust tests for production quantum applications.<h3>Highlights</h3><ul><li>Deterministic Debugging: Step through instructions with full statevector snapshots, classical bit snapshots, and support for control-flow constructs such as IfElseOp, loops, and SwitchCaseOp.</li><li>Prefix Analytics: Compute exact probabilities, sampler-backed probabilities, counts, marginals, expectation values, Shannon entropies, divergences, and total-variation distances for every prefix of a circuit.</li><li>Testing Confidence: Assertion helpers compare counts, probabilities, and statevectors with configurable tolerances for noisy hardware.</li><li>Operational Insights: CSV/JSON export, pandas integration, ASCII/Matplotlib histograms, and structured trace logging make it simple to observe and share what your primitives are doing.</li></ul>Qiskit Inspect is open source and licensed under the Apache License 2.0, making it a valuable resource for developers working with quantum circuits.

Qiskit Inspect

<h3>A Lightweight, High-Performance Mathematics &amp; Quantitative Finance Toolkit</h3>DervFlow is a comprehensive toolkit designed for high-performance mathematics and quantitative finance applications. It offers a range of features including production-ready option pricers, risk analytics, portfolio construction utilities, time-series diagnostics, and yield-curve analytics. These capabilities are backed by rigorously tested numerical kernels implemented in Rust for enhanced safety and speed.Highlights:<ul><li>Options Pricing: Includes Black-Scholes-Merton analytics, implied volatility solvers, and exotic payoffs exposed through ergonomic Python classes.</li><li>Risk Analytics: Offers portfolio Greeks aggregation, historical/parametric/Monte Carlo VaR and CVaR, drawdown analytics, and performance ratios.</li><li>Portfolio Construction: Provides mean-variance optimization, efficient frontiers, and Black–Litterman blending of market and investor views.</li><li>Yield Curves: Features bootstrapping from bonds or swaps, Nelson–Siegel parametrisations, and bond analytics.</li><li>Time Series: Supports return transformations, rolling statistics, autocorrelation diagnostics, and GARCH-family volatility models.</li><li>Monte Carlo Simulation: High-throughput stochastic process simulators with optional parallel path generation.</li></ul>

DervFlow

Minitensor is a lightweight, high-performance tensor operations and automatic differentiation library designed for developers seeking a PyTorch-like experience with enhanced speed and efficiency. Built with a Rust backend, it ensures maximum speed and memory efficiency, while its Python bindings provide a familiar API for easy adoption. Minitensor supports complete neural network layers and optimizers, offering seamless interoperability with NumPy arrays. Key features include automatic differentiation for gradient computation, a modular design for customization, and a high-performance engine optimized with SIMD. The library is currently in active development, making it a cutting-edge choice for developers looking to build and deploy intelligent applications.

MiniTensor

Interactive semantic maps (WordNet) via CLI and web. WordAtlas builds interactive semantic maps from <a target="_blank" rel="nofollow" class="break-all link-break text-gray-500 hover:underline" href="https://wordnet.princeton.edu/">WordNet</a>.

WordAtlas

Test your reflexes and number recognition skills by popping the right bubbles before time runs out!

Bubble Game

Easily design, simulate, and visualize quantum circuits with intuitive tools that allow you to experiment with quantum gates and qubit states effortlessly.

QSim - Quantum Simulator

A small Rust CLI/library that demonstrates RSA primitives for learning, and provides secure flows using OAEP (encryption), PSS (signing), and a hybrid RSA+AES-GCM scheme for large files.