Hi, I’m Ritik Yadav, an SDE at ClearMind (Ex-AlphaQuark). I specialize in engineering high-throughput real-time trading engines, WebSocket pipelines, mobile FinTech applications, and AI memory & agent platforms like MemLite and Raspond.
Computer Science graduate with demonstrated success in architecting production backend services, mobile trading apps, and AI platforms. Strong algorithmic foundation paired with low-latency system design.
Proven expertise in cutting execution latency down from 500ms to <40ms under peak trading loads using Redis caching, asynchronous pub/sub pipelines, and optimized PostgreSQL query indexes.
Architect of agentic workflows at Raspond (intent parsing and document ingestion) and creator of MemLite, a local-first memory engine for LLMs with FAISS semantic retrieval.
Shipped 3 live commercial investment and trading applications to Google Play Store (ARFS, AlphaPro, Zamzam Capital), serving over 2,000+ active traders with live market feeds.
Interactive engineering expedition tracing systems architecting, mobile product scaling, and algorithmic foundations.
Core backend engine fixes, retrieval bug fixes, and infrastructure contributions merged and active across leading open-source AI and developer ecosystems.
Leading open-source RAG engine based on deep document understanding. Contributed core retrieval optimizations, layout analysis fixes, and parser chunking improvements.
tbls in document book chunking algorithms.
Industry-standard application performance monitoring and error tracking platform. Contributed backend validation for cron monitor schedules.
src/sentry/monitors/validators.py to prevent erroneous schedule rejection.
High-throughput Universal I/O Proxy & Load Balancer for 100+ LLMs in standard OpenAI format. Contributed streaming hooks and router rate limiting.
Original developer frameworks, AI memory engines, and production systems with measurable architectural impact.
Local-first memory decay and semantic ranking engine for LLMs. Implements FAISS vector indexing, exponential recency decay, and SQLite persistence. Includes a custom Open VSX extension rendering interactive AI thought graphs and synchronizing workspace context.
AI-driven conversational workflow platform that parses user intent and orchestrates dynamic agent actions. Built the multi-format document ingestion pipeline and integrated Langfuse for end-to-end LLM observability.
Live trading applications published on Google Play Store featuring low-latency market feeds, multi-channel notification engine (WhatsApp/Telegram), and dynamic advisor discount modules.
Accessibility platform translating spoken audio and text into synchronized 3D sign language animations to bridge communication barriers for speech- and hearing-impaired communities.
Computer vision surveillance solution calculating vehicle density in real-time from camera feeds to dynamically balance traffic signal timers and flag perimeter security violations.
Full-duplex web messaging application paired with a multi-user collaborative drawing canvas using HTML5 Canvas primitives and WebSocket broadcast channels.
Core engineering competencies, languages, frameworks, and architecture patterns.
Academic foundation, competitive programming stats, publications, and certified honors.
Graduated August 2025 • Kanpur, India
Ranked in the top 5th percentile nationwide in the Adobe GenSolve challenge for algorithmic problem solving. [View Certificate]
Solved over 700+ algorithmic problems across LeetCode (500+) and GeeksforGeeks (200+) focusing on dynamic programming, graphs, and low-overhead algorithms.
Author of deep-dive articles including "How GPT Works" on Medium and "URL Listening in React Native WebView".
Earned official skill credentials in Google Cloud Platform infrastructure and Generative AI services. [Verified Google Profile]
Available for SDE, Backend Engineering, and Distributed Systems roles. Reach out directly via email or connect on LinkedIn.