Dev Joshi Dublin 18, Ireland | +353858088571 | devdeepak157@gmail.com | LinkedIn | GitHub | LeetCode | Portfolio Education Trinity College Dublin Dublin, Ireland B.A. (Mod.) in Computer Science 2024 – 2028 • Relevant Coursework: Data Structures & Algorithms, Programming (Java), Systems Programming (C), Discrete Mathematics, Statistical Analysis, Applied Probability Experience Junior Full-Stack Developer Sep 2023 – Feb 2024 ArrowGlue (Ed-Tech Startup) Remote – Sedona, US • Improved user authentication by 60% , by integrating OAuth2-based social login (Google, GitHub) and caching tokens with hash map lookups. • Reduced query response time by ∼ 30% by restructuring PostgreSQL queries with select related and B-tree indexing when a user searches their saved diagram (O(log n) vs O(n)). • Improved sprint deliverable completion rate by 20% by running weekly Google Meet calls and helping the founder prioritise the backlog. Projects Emerald Health – AI Healthcare Assistant [GitHub] | [Live Demo] Jun 2025 Google Student AI Hackathon 2025 — React, TypeScript, Firebase, Gemini API • Built a multimodal AI healthcare assistant MVP, in under 5 hours , by designing backend LLM pipelines and implementing efficient queue-based task scheduling for symptom and image processing with a team of 4. • Generated medical summaries for GP review, with 100% test case accuracy , by building a hash-map-backed vision-text pipeline using the Gemini API to store and retrieve intermediate results efficiently. • Automated GP booking workflows, as measured by a 90% reduction in manual booking time , by implementing priority queues and drafting emails to local GP while integrating real-time clinic geolocation. PolicyPal – AI Insurance Assistant [Live Demo] Sep 2025 TechIreland National AI Challenge 2025 — Python, GPT-4/3.5, Flask, React • Built a tool that summarises and transforms dense insurance PDFs into interactive dashboards, delivering insights in under 90 seconds vs. 2+ hours manually • Extracted and categorised coverage while automatically flagging exclusions, using parsing and rule-based recognition. • Cross-referenced policy data with real-time datasets (flood zones, crime stats) to generate actionable recommendations. AlgoPlayground – Interactive DSA Visualiser [GitHub] | [Live Demo] Ongoing Personal Project — React, TypeScript, Node.js • Building a tool to help students in CSU22011 – Algorithms and Data Structures I at Trinity College Dublin better understand core data structures & algorithms. • Developing interactive visualisations for stacks, queues, linked lists, and hash tables to make abstract concepts easier to grasp. • Aiming to reduce the barrier of entry for classmates by transforming dense theory into interactive and fun learning. Technical Skills Languages : Java, Python, C, JavaScript, HTML/CSS, R Frameworks : React, Node.js, Django, Flask, Tensorflow, FastAPI Developer Tools : Firebase, Git, Docker (learning), Google Cloud Platform Libraries : pandas, NumPy, Matplotlib, Tensorflow AI Tools : Gemini API, OpenAI API, LangChain, Hugging Face Transformers Leadership / Activities S2S Mentor, Trinity College Dublin : Mentoring a group of 13 first-year Computer Science students , providing academic guidance, career advice, and support in transitioning to university life. Strategic Planning Lead – S2S Summer Committee : Driving the development of a 5-year strategic plan by leading a team of 4, coordinating weekly calls, and aligning initiatives with the S2S coordinator to ensure actionable outcomes. Google Student AI Hackathon 2025 Participant : Developed Emerald Health AI assistant MVP, gaining hands-on experience in AI integration, teamwork, and rapid prototyping.