Suraj Barman Rewriting a Python RAG Library in Rust: Context, Implementation, and Best Practices Context History of Python RAG Libraries Retrieval‑augmented generation (RAG) emerged as a practical way to combine large language models with external knowledge sources. Early open‑source implementati...
Suraj Barman Blitz’s Rust‑Powered NoSQL Backend Revamp for League of Legends Blitz’s Rust‑Powered NoSQL Backend for League of Legends Blitz has rebuilt the core data layer of its League of Legends service using a Rust rewrite and a NoSQL store, abandoning the legacy PostgreSQL...
Suraj Barman Protect AI Agents from Malicious Skills – Clean Code & TDD Practices Protecting AI agents from malicious skills requires clear boundaries, disciplined code, and automated verification. Identify Potential Threat Vectors Understanding where hostile code can enter an AI s...
Suraj Barman Claude Opus 4.6 vs GPT‑5.3 Codex: Technical Guide to AI‑Driven Software Engineering Context History AI‑assisted programming has moved from experimental prototypes to production‑grade tools in just a few years. Large language models now generate code, suggest fixes, and even write tes...
Suraj Barman QIELend: Efficient DeFi Lending on the QIE Blockchain – Technical Overview QIELend: Efficient DeFi Lending on the QIE blockchain QIELend is a decentralized lending protocol built on the QIE blockchain, aiming to lower transaction costs and improve capital efficiency for borr...
Suraj Barman Understanding Human Identity When Tortured: A Poetic Guide What does it mean to be human when subjected to torture ? This question frames Sam's poem as a mirror for the reader, inviting a reflective stance on dignity, pain, and agency. The work blends persona...
Suraj Barman Sai Perps Platform: Merging CEX Speed with On‑Chain Settlement Context & History Sai entered the crypto market in 2021, focusing on transparent communication tools for blockchain projects. In early 2024 the company identified a gap: traders demanded the instant o...
Suraj Barman Zero‑Copy C++ ONNX Inference for Low‑Latency Video: Achieving 29 FPS Zero‑Copy C++ ONNX Inference for Low‑Latency Video The ONNX runtime, when combined with native C++ bindings and zero‑copy memory techniques, enables real‑time inference on video streams with minimal l...
Suraj Barman How Santander Cut Provisioning Time from 90 Days to Hours with the Catalyst Platform Santander’s Catalyst platform abstracts cloud infrastructure, enforces architectural standards, and accelerates service delivery for billions of daily transactions. Core Architecture The platform cent...
Suraj Barman Effective Market Writing: Cutting Through Noise with Structure and Timing # Context History The practice of market writing has often been cluttered with vague statements and overused buzzwords. Historically, analysts tried to impress with complex jargon rather than deliveri...
Suraj Barman HackerNoon Trending Tech Stories Ranking Methodology and 2026 Top Picks HackerNoon Trending Tech Stories Ranking HackerNoon ranks its technology stories by combining three core signals: total pageviews , measured user engagement , and the volume of comments. Each signal i...
Suraj Barman How Dataops.live Automates Trusted AI at Scale Dataops.live provides an end‑to‑end DataOps automation platform that embeds trustworthy AI throughout every stage of the data lifecycle. Automated Data Ingestion Dataops.live connects to diverse sourc...