Feb 2025 – Present · 1 yr 3 mos · Bengaluru, Karnataka, India · On-site

• As part of Adobe Illustrator’s Rendering team, I work at the intersection of GPU systems engineering and Generative AI infrastructure, building high-performance pipelines that bring intelligence and efficiency to creative applications.
• Agentic Framework for Illustrator Effects: Architected a plug-and-play agentic system that translates natural language prompts into fully editable Illustrator effects. Integrated context-aware conversational interfaces, state tracking, and client-side validation to ensure robust responses from LLMs (GPT-4, Gemini, Claude) and VLMs (BLIP-2, Kosmos-2) — cutting effect discovery and application time by 50%.
• Codebase Indexing using Agentic RAG: Led the design of a retrieval-augmented generation (RAG) system that indexes internal codebases into a vector database, enabling semantic, context-aware search and cutting code discovery time by 30%.
• Transform Operations Optimization: Enhanced the Illustrator transform pipeline through GPU tessellation, caching, and parallel compute scheduling, improving frame rates by 5–12× for over 87% of users.
• These projects combine AI integration, GPU performance, and system reliability, reflecting my focus on building scalable, production-grade AI infrastructure for creative tools.
• Core Skills: GPU Programming · C++ · Python · Agentic AI Systems · LLM Orchestration · Retrieval-Augmented Generation · PyTorch · Vector Databases · System Optimization · Multimodal AI Integration

Jul 2023 – Feb 2025 · 1 yr 7 mos · Bengaluru, Karnataka, India · On-site

• Worked on performance, authenticity, and AI reliability initiatives in Adobe Illustrator’s rendering and intelligence systems.
• VerifiAI: Engineered a hybrid ML pipeline for detecting AI-generated and modified media using forensic noise analysis and Diffusion Reconstruction Error-based classifiers. Achieved 95% accuracy while aligning with C2PA and Adobe’s Content Authenticity Initiative to create a verifiable trust layer for generative content.
• Telemetry Logging & Insights: Built an in-memory telemetry pipeline capturing 15,000+ events/day without runtime overhead. Integrated with Databricks for large-scale analysis, surfacing performance trends that improved Illustrator’s 60 FPS stability from 32% to 70%.
• Infinite Resolution Textures: Adapted Nvidia’s Infinite Resolution Textures research to Illustrator’s rendering engine, enabling super-resolution rendering and zoom operations for large canvases.
• These projects strengthened my expertise in machine learning pipelines, system telemetry, and GPU optimization, as well as my ability to operationalize research for production-scale creative software.
• Core Skills: ML Infrastructure · GPU Optimization · Data Engineering · Diffusion Models · PyTorch · Databricks · C++ · AI Reliability · Performance Analytics