Taniya Souza
Software engineer
ABOUT
I'm an AI Product Engineer and builder who loves turning complex models into lightning-fast, user-friendly products. Whether I'm building data pipelines that analyze GitHub at scale or optimizing LLM latency for applications, my focus is bridging the gap between cutting-edge AI and practical reality.Frontend.
ReactJS
HTML & CSS
JavaScript
TypeScript
Tailwind CSS
Storybook
Figma
Backend.
NodeJS
ExpressJS
MongoDB
AWS
Supabase
Qdrant
REST APIs
LLM/AI integrations
CI/CD
MySQL
Languages & Tools.
Python
Next.js
SQL
Linux
ESLint/Husky
EXPERIENCE
AI Foundations
March 2026 – Present
TypeScript
Performance
API Integration
Frontend Engineer (TypeScript, Performance & API Integration)
Building GitPullTalent: a GitHub-based developer ranking and discovery platform. Designed a scoring system evaluating developers on commits, pull requests, and repo quality for recruiter-focused signals. Handled large-scale data ingestion, ranking logic, and profile prioritization heuristics.
Indhic AI
Jan 2026 – Feb 2026
AI Engineer
Consultant
Contract
AI Engineer Consultant (Contract), Bengaluru
Comparative analysis of healthcare LLMs (MedGemma vs Gemini Flash 2.5 & 3.0) across clinical reasoning, cost, latency, and deployment. Extracted treatment-planning algorithms from STOPP/START frameworks.
Grapevine (Series-A)
July 2025 – Jan 2026
AI Product Engineer
Intern
AI Product Engineer Intern, Bengaluru
Deployed LLM-powered conversational and voice AI using Gemini, ElevenLabs, and Google Voice. Migrated n8n workflows to Python, reducing execution time by ~97%. Built embedding-based resume tagging and search pipelines using Qdrant. Built backend automation for job discovery and outreach at scale.
frog by Capgemini
January 2025 - June 2025
ReactJS
TypeScript
Next.js
Frontend / Design technologist
Automated insurance policy verification using n8n and DeepOpinion. Built reusable, accessible UI components with React, TypeScript, Tailwind, and Storybook. Delivered a full-stack journaling app with Supabase and enforced coding standards via ESLint/Husky.
PROJECTS
GitPullTalent
- Goal: Streamline tech hiring by replacing subjective resumes with quantitative GitHub activity metrics.
- Impact: Engineered a robust ranking system evaluating commit quality, PR frequency, and repo health.
- Tech: TypeScript, Next.js, BullMQ.
github-data-pipeline
- Goal: Build a headless data pipeline for processing and analyzing large-scale GitHub repository data.
- Impact: Designed scalable ingestion and caching strategies, handling real-time rate limiting and large payloads.
- Tech: TypeScript, BullMQ, Drizzle ORM, Postgres, Redis.
Fitness App
- Goal: Monitor user posture and movements in real-time to provide automated physical feedback.
- Impact: Implemented AI-driven posture detection, integrating seamless communication between the frontend and the AI server.
- Tech: OpenCV, Python, React, AWS.
JournalMind
- Goal: Create a full-stack, aesthetically pleasing progressive web application for journaling.
- Impact: Architected a robust, minimal design system with instantaneous interactions leveraging SSR.
- Tech: Next.js, TypeScript, Tailwind CSS, Supabase.
Image Processing AI
- Goal: Automate complex image editing tasks like precise object removal and generative fill.
- Impact: Orchestrated three state-of-the-art vision models to pipeline a localized inpainting system in Python.
- Tech: Grounding DINO, Segment Anything (SAM), Stable Diffusion XL.
Positivus
- Goal: Develop an award-winning layout for a digital marketing agency interface.
- Impact: Achieved a highly expressive design with buttery smooth micro-interactions that elevate brand perception.
- Tech: ReactJS, Tailwind CSS, GSAP.
Travel Planner
- Goal: Provide a unified hub for tourists to generate, map, and organize smart itineraries dynamically.
- Impact: Integrated live data sources to accurately assemble travel timelines.
- Tech: React, APIs, Maps Integration.
