Résumé

Multi-agent systems, LLM APIs, retrieval (RAG), Python · team of 2 interns, reporting to a manager and an associate

• Building an AI marketing-strategist system for Sakal’s sub-brands and client brands, feeding into live campaign workflows.
• Designed a multi-agent LLM panel that debates and evaluates campaign strategies, grounded in each brand’s documents.
• Generated platform-specific social content suggestions: what to post and when.

HTML, CSS, JavaScript, Figma, ScrollReveal.js, GoDaddy

studio-rangeen.art

• Designed, developed and deployed the art studio’s whole website.
• Worked with 3 other designers and 1 other developer.
• Maintained the site for another 2 months.

TypeScript, offline-first design, Marathi vocabulary research · team of 4

anikadam.com/work/chetana-aac

• Leading a free, offline Marathi communication app for ~50 children with little or no speech.
• Drafted a 193-concept core vocabulary and 12 written design decisions; built a browser-based board design tool.

Embedded C, STM32F411, ESP32-S3, I²S/DMA, Python · team of 6

anikadam.com/work/sih26

• Microphone capture streaming over USB on two boards; mic_lab.py, a capture and test tool that works with any board.
• On-device enhancement target: speech more than 15 dB clearer within 40 ms.

Python, OpenCV, MediaPipe, signal processing · built at Ignisia, MIT-WPU’s national AI hackathon

github.com/aryamaddel/CardioVision-rPPG

• Built a real-time heart-rate pipeline from face video using MediaPipe tracking and the POS algorithm.
• Added Eulerian video magnification to show blood-flow colour changes next to the BPM readout.
• Re-scoped the project overnight after a surprise mid-hackathon change and shipped within the 36-hour window.

Python, YOLOv8, OpenCV, Kaggle, Google Colab

• Trained a YOLOv8 model on a public damage-detection image dataset to detect and locate damage.
• Trained and tested on Kaggle, then ran the pipeline on Google Colab.

Arduino Uno, AWS IoT Core, DynamoDB, React/Vite

• Built a pipeline from piezoelectric footstep sensors through AWS IoT Core to a DynamoDB-backed React dashboard.
• Documented the full system architecture in a formal report.

• CGPA 8.03 after semester 6