Headline

Software Engineer @ Google | Gen-AI | Lead | Backend | Architect | WSE | Full-Stack | Polyglot | ML | RAG | Stanford | DU

About

Hi I am a backend-first full-stack AI engineer and system architect with 7+ years of experience building and scaling products across early-stage startups and mature, high-traffic platforms. I specialize in designing complex backend systems and AI-native architectures, with deep hands-on experience in RAG-based applications and data pipelines. I build end-to-end systems—from document ingestion and indexing to embeddings, vector search, retrieval strategies, orchestration, and production deployment. What sets me apart is adaptability. I’m comfortable switching tech stacks on demand—languages, databases, cloud providers, or architectural patterns—based on system constraints and business goals. Tools are interchangeable; outcomes are not. My work spans: Distributed systems, microservices, and high-throughput APIs Real-time and event-driven architectures Cloud-native design on AWS, Azure and GCP AI pipelines using LangChain/LangGraph, vector databases, and multiple GenAI providers I’ve operated in environments where speed and ownership matter just as much as stability and scale, which allows me to ship fast without accumulating avoidable technical debt. I enjoy building systems that are cleanly architected, scalable by design, and resilient in the real world—especially when backend complexity and AI meet. If you’re building something ambitious and need someone who can own backend architecture, adapt quickly, and think long-term, I’m ready. 👉 Always eager to learn, adapt, and contribute to impactful products.

Location

Bengaluru

AI stack

antigravity

claude code

claude

codex

cursor

github copilot

Work experience

Software Engineer(WSE) at Google, Jul 2026 - Present · 4 mos, Bengaluru, Karnataka, India, As a Web Solutions Engineer within the Digital Support Experience team, I design, build, and scale state-of-the-art AI agents and intelligent automation tools that optimize operations and transform the Google Cloud support experience. Leveraging Google's world-class engineering infrastructure, I build full-stack solutions that convert complex supportability strategies into seamless, production-grade applications. Key Responsibilities & Impact 🤖 AI Agent Development: Architect and deploy intelligent, context-aware AI agents utilizing Google’s internal LLM frameworks and Gemini APIs to automate complex support workflows and improve resolution times. ⚙️ Automation & Tooling: Design and develop robust, scalable full-stack automation tools and internal supportability platforms that drive operational efficiency across Google Cloud Support. 🛠️ Google Engineering Ecosystem: Build end-to-end solutions using Google Cloud Platform (GCP), modern JavaScript/Python frameworks, APIs, and cloud-native databases. 🤝 Cross-Functional Collaboration: Partner closely with Supportability, Strategy, and Engineering teams to translate complex business requirements into elegant technical solutions. 📈 Scalability & Supportability: Focus on building highly reliable and reusable software systems, establishing best practices for code quality, testing, and continuous deployment. Core Technologies AI/ML: Generative AI, AI Agents, Prompt Engineering, Large Language Models (LLMs), Vertex AI Software Engineering: Full-Stack Development, Python, JavaScript/TypeScript, API Design, System Architecture Cloud & Infrastructure: Google Cloud Platform (GCP), Cloud Functions, App Engine, AppSheet, Automation Tools Methodologies: Agile, CI/CD, Supportability Engineering, System Integration

Technical Lead at The Carlyle Group, Feb 2026 - Jul 2026 · 6 mos, Gurugram, Haryana, India, Architected and implemented a benchmark-driven pricing engine for financial facilities, integrating Snowflake (market analytics) with Postgres (composition + configuration) to automate market-based yield calculation at scale. Designed a cross-system data architecture enabling weighted benchmark construction, time-versioned composition snapshots, and full audit trails for financial valuations. Built a scalable Snowflake service layer (Python) with stored-procedure based writes and view-based reads, ensuring idempotent bulk ingestion, high-performance analytics queries, and fault-tolerant connectivity. Developed a Benchmark Rate Resolver service to compute dynamic discount rates using weighted average yield, illiquidity premium, and risk adjustments — with caching, batch resolution, and staleness detection for high-volume valuation runs. Engineered time-series analytics aggregation capable of handling benchmark rebalancing across date ranges, ensuring accurate weighted metrics, coverage calculations, and composition transition tracking. Integrated AI-assisted validation and anomaly detection for historical yield uploads, identifying sparse data, coverage gaps, and inconsistencies before valuation execution. Preserved backward compatibility across legacy pricing logic by decoupling discount rate resolution from cashflow modeling (standard vs bloomberg modes), ensuring zero regression in existing deal executions. Implemented structured audit capture for every valuation run, including composition snapshot, spread movements, yield decomposition, and data freshness indicators.

Senior Software Engineer at Incyte, Aug 2024 - Feb 2026 · 1 yr 7 mos, Bengaluru, Karnataka, India, Designed and developed AI agents for drug discovery applications, leveraging RAG (Retrieval-Augmented Generation) for intelligent document and media processing with WebSocket and WebRTC. Built and optimised AI-powered search and retrieval systems using LangGraph, LangChain, OpenSearch (vector database), and DynamoDB. Developed end-to-end pipelines for document ingestion, indexing, and querying, integrating AWS S3, OpenSearch, and DynamoDB for scalable storage and retrieval. Ensured MLR compliance by implementing AI-driven claim validation and regulatory checks against approved medical content. Worked with AWS and Azure cloud ecosystems, optimising the deployment and performance of AI-driven applications and integration with different GenAI Models like OpenAI, Gemini, Claude, e.tc.

Software Engineer at Docquity, Aug 2022 - May 2024 · 1 yr 10 mos, Gurugram, Haryana, India, A community of experts and leaders in health tech. Exchange ideas with the best doctors in the region. Receive educational resources designed by well-known and respected leaders. Attend virtual discussions with world-class speakers. At Docquity, I worked as a Back-end engineer to re-engineer an existing monolithic architecture into a microservices-based architecture. I introduced gRPC for inter-service communication and added caching for fast data retrieval. I wrote several services for features like community, webinars, content management, etc., which handle thousands of concurrent users and high-volume, real-time data processing. Content Service - This service works as a storehouse for all other services; it manages all types of content like text, audio, video, etc. Tech stack - Springboot(JAVA), Python, MongoDB, Redis, gRPC, Jenkins, Docker, Websocket Community Service—This is a feature service where doctors can create a community according to their speciality and actions—join, invite, exit, block. The tech stack is Python(Django), MySQL, and AWS, Nest.js(Node), Python, PHP, MySQL, AWS.

Software Engineer at Zenarate, Jun 2021 - Jul 2022 · 1 yr 2 mos, Gurugram, Haryana, India, Headquartered in Silicon Valley, Zenarate's AI Coach is used by top 10 U.S. financial institutions, the largest BPOs, and leading companies in the healthcare, technology, and services industries. Zenarate AI Coach clients experience 30-60% KPI lift creating top-performing customer-facing teams through conversation simulation. At Zenarate, I worked as a Full-Stack-Developer to add new features to existing applications. I added WCAG compatibility(accessibility) to the application, i.e., the website can be used by visually impaired people with the help of tools like VoiceOver (MAC) and JAWS(Windows). Tech Stack - JavaScript, jQuery, React.js Added a scheduled assignment feature to create Stories of AI chat and audio sessions to boost the KPI of financial organisations by 30%. Tech stack - Python, Node.js, JavaScript, MySQL Added Spell-check and grammar-check features to the existing AI chatbot. Tech stack - Python(Flask), Node.js, JavaScript, MySQL, integration with different GenAI Models like OpenAI

Education

Master's degree, Informatics, Delhi University

Online Credit Course, Machine Learning, Stanford University Department of Computer Science

Public links

github: https://github.com/avinasx

avinasx.github.io: https://avinasx.github.io/