
Full Stack and Health AI Engineer Intern
Embargo Media · Seoul, South Korea
1. Developed a multilingual hearing and cognitive screening platform, enabling end-to-end digital assessments for 50 pilot participants across 8 languages. 2. Implemented an interactive cognitive screening workflow with secure data handling and privacy-conscious image processing, validated across 60 test sessions. The platform provides non-diagnostic screening outputs to support preliminary assessment. 3. Developed a location-aware clinic discovery feature that enables users to identify nearby healthcare providers, with fallback functionality for limited location access and responsive search performance across 30 clinics.
Trainee Engineer
Eximietas Design · Bangalore
1. Architected an AI-driven workflow generator processing 1,000+ workflows monthly, converting single-prompt inputs into structured multi-step workflows and reducing manual configuration time by 60 % using Fast API, Gemini API, Next.js, React, and TypeScript. 2. Engineered an agentic code-generation pipeline with CrewAI, LangChain, and OpenAI APIs to auto-generate production-ready Kotlin Android modules, reducing prototype development time by 85 % (from 4 hours to 30 minutes) and accelerating sprint delivery. 3. Designed scalable backend infrastructure with RESTful API routing, JWT-based authentication, and vector-based reasoning using Pinecone to support complex workflow orchestration for 50+ concurrent users.
LLM Researcher
Algoverse AI Research · Palo Alto, California, United States
1. Investigated the Context Rot Problem in Vision-Language Models like Qwen-2.5-VL by building a Captioning pipeline using the GPT-4o API with caption lengths of about 50 to 500+ words. 2. The Caption-Augmented inputs caused an 8% accuracy degradation when benchmarking Qwen-2.5-7B and GPT-4o on the OKVQA dataset.
Android Development Intern
Open MedLabs · Bangalore
1. Engineered emergency alert features using AWS SNS/SES delivering notifications to 500+ users, improving delivery reliability to 99.2 % and reducing user onboarding time by 40 % via Firebase Authentication and real-time database integration. 2. Developed real-time patient monitoring dashboards handling 10,000+ data points daily with sub-1s update latency using Jetpack Compose, AWS RDS, and WebSocket connections for live telemetry streaming. 3. Built modular UI components and adapters using RecyclerView, Intents, and MVVM architecture, improving code maintainability by 50 % and reducing UI rendering time by 30 % across 15+ app screens.

KNOTiC
9281 members

AI Mobile Coders
1631 members


