Apr 2025 – Mar 2026 · 11 mos · Gurugram, Haryana, India

• Kubernetes Sticky Pods Issue→ Diagnosed and resolved critical sticky client-side connections problem across Magicpin's internal services, reducing unnecessary autoscaling and cutting cloud costs by while significantly improving system reliability.
• Whatsapp Messaging Service Migration → Led the migration of the core Whatsapp messaging system from Gupshup to Karix, reducing cost and manual effort by automating the transfer of 2000+ templates while making the system fully backwards compatible, ensuring zero service disruption during the transition of high-volume production traffic.
• Security & Compliance → Mitigated a high-severity HTML Injection vulnerability by implementing robust server-side sanitization, effectively securing the platform against XSS attacks.
• API Performance Optimization → Optimised high-traffic composite APIs, reducing response sizes by over 60% and significantly lowering CPU utilization by decoupling heavy data dependencies, saving CPU and memory footprint by 40%
• Scheduled Offer Feature → E2E implementation of catalog scheduled offer feature which reduced manual offer operation time by ~90%

Nov 2024 – Apr 2025 · 5 mos · Gurugram, Haryana, India

• Vector Search Architecture → Architected and deployed Vector Semantic Search (prod search DB) for 50M+ documents from scratch. Tuned HNSW parameters and reduced vector dimensions (3072 → 512), cutting latency by 66% (750ms → 250ms) and storage costs by 83% while improving semantic relevance.
• MLT Service → Built the "More Like This" widget API using FastAPI and Aerospike. Engineered a custom scoring algorithm and optimized caching strategies to increase cache hit rate (46% → 80%) and reduce p99 latency (320ms → 60ms).
• GenAI & LLM Pipelines → Designed an end-to-end Fashion LLM Tagging pipeline. Exposed endpoints for bulk CSV ingestion and utilized OpenAI APIs to automate attribute generation, improving search recall and reducing manual tagging effort by 90%.
• Search Backend Ownership → Took end-to-end ownership of critical search workflows, delivering production-grade features. Collaborated extensively across cross-functional teams to integrate complex search requirements into scalable microservices.

May 2024 – Jul 2024 · 2 mos

Jul 2023 – Jan 2024 · 6 mos

• • Achieved 100% organic growth for event business, reaching ₹3.5 million revenue with zero marketing spend and no upfront investment.

Jul 2022 – Jul 2023 · 1 yr

• • Led a 40-member team to organize college trips and events for 500+ participants, overseeing logistics, budgeting, and event execution.

Jun 2022 – Jun 2024 · 2 yrs · Delhi, India

• Artificial Intelligence Research Group - Netaji Subhas University of Technology is a student-run, Non-profit community of Data Science and Machine Learning enthusiasts, backed by Intel®.
• We conduct research and teach a wide range of AI-related topics, often cross-disciplinary. We collaborate internationally and locally with academics, businesses and public stakeholders, working at the frontiers of knowledge, solving real-world problems, and in policy and public outreach.

Dec 2020 – Jan 2021 · 1 mo

• Worked on a project under the mentorship of the IBM US/India team to solve a real-world problem using ML.

Jan 2020 – Jan 2021 · 1 yr

Jan 2019 – Jan 2020 · 1 yr

Jan 2018 – Jan 2019 · 1 yr