Siddhant Parekh

Software engineer with 5+ years building scalable, AI-powered products end-to-end. Tech lead currently driving multiple parallel engineering teams on a customer engagement platform serving 100K+ active users and processing 2M+ daily messages. Specialized in backend systems, LLM and RAG-based applications and ML pipeline orchestration across Node.js, AWS and Azure AI stacks.

GYTWorkz: Software Development Engineer - II

• Lead end-to-end delivery of Engage, a customer engagement platform serving 20+ enterprise clients, 100,000+ active users and 1M+ customers; own architecture, sprint planning, solution design and hands-on full-stack development.

• Manage multiple parallel feature pods (3–4 engineers each, totalling 10 engineers at peak), driving technical mentorship, code reviews and cross-team coordination to ship continuous releases.
• Spearheaded a communications module processing 2M+ daily SMS and email messages across client applications, enabling real-time customer interaction at scale.
• Re-engineered the SMS service, cutting throttling by 90% and achieving up to 5x faster delivery, ensuring reliability for the 2M+ daily message workload.
• Architected a smart scheduling module that reduced meeting coordination time by 40% through automated optimal-time suggestions.
• Contributed to Onyx, an in-house AI application framework powering LLM-driven assistants and retrieval-augmented (RAG) workflows; built backend integrations, prompt orchestration and Azure AI Foundry deployments for client-facing AI features.
• Represented Onyx at GITEX Dubai 2025, leading technical demos and architecture conversations that converted into 2 new enterprise clients for the AI platform.
• Introduced standardized task management, deployment and QA processes, reducing client-reported bugs by 50% and tightening release cycles.

Stack: Node.js, Express, PostgreSQL, MongoDB, React, AWS (ECS, Lambda, RDS, S3), Azure AI Foundry

• Built a blockchain-based supply chain management platform for the pharmaceutical industry, enabling end-to-end traceability, tamper-proof records and regulatory compliance across distributors and manufacturers.

• Architected a MongoDB indexing layer for blockchain events, optimizing data retrieval by 70% by eliminating redundant RPC calls and enabling complex filtered queries that are natively unsupported by the chain.
• Designed and shipped an LLM-powered WhatsApp chatbot for real-time supply chain tracking, integrating the WhatsApp Business API with backend services using a retrieval-and-summarization pattern to fetch, format and surface shipment data through natural-language interactions.
• Implemented role-based access controls and privacy guardrails within the chatbot workflow, securing sensitive supply chain data while meeting regulatory and data-privacy standards.
• Migrated the frontend data layer from REST fetch patterns to RTK Query, improving state management, reducing boilerplate and lifting UI performance.

Stack: Node.js, MongoDB, React, Redux Toolkit / RTK Query, WhatsApp Business API, Hyperledger

• Contributed to Bhashini, the Government of India’s national language translation initiative, as part of the core team responsible for aggregating ML models from research labs across India and building end-to-end multilingual video translation pipelines for NPTEL to improve course accessibility for regional learners.

• Architected a modular ML pipeline combining Speech-to-Text, Machine Translation and Text-to-Speech models, dynamically selecting optimal model combinations based on source and target language constraints across multiple Indian languages.
• Built configurable translation workflows automating video ingestion, transcription, translation and audio synthesis for the NPTEL course catalogue, reducing manual processing effort by 40%.
• Integrated and benchmarked ML models from different research labs into a unified inference layer, improving overall translation accuracy by 20% through intelligent model orchestration and fallback logic.

Stack: Python, Node.js, ML pipelines (STT / MT / TTS), Docker, REST APIs