About

Graduate Student Researcher at the UC Davis Alzheimer’s Disease Research Center, building the I‑Care platform that enables remote caregivers to support older adults with Alzheimer’s, including real‑time chat, video calls (WebRTC), adaptive reminders, and a personalized task calendar. Previously, full‑stack Software Engineer at Capgemini for 2+ years, serving the global bank ABN AMRO on digitization of commercial banking process.

GenAI & LLMs Multimodal ML React / Vue Firebase Java / SpringBoot Azure • AZ‑900

Experience

Graduate Student Researcher — UC Davis Alzheimer’s Disease Research Centre

Mar 2025 – Present · React, VueJS, JavaScript, Firebase, Python, PyTorch
  • Building and enhancing I‑Care to enable remote caregiver support with chat, WebRTC video calls, adaptive reminders, and a personalized task calendar.
  • Optimized Firebase backend across tasks/goals/events/access controls, reducing redundant reads/writes and improving responsiveness by 30%.
  • Co‑developing a GenAI‑augmented system using speech, motion, and vitals data to detect cognitive decline and trigger adaptive interventions.

Software Engineer — Capgemini

Aug 2021 – Aug 2023 · Java, JavaScript, VueJS, Git, NPM, REST
  • Full‑stack developer for ABN AMRO, digitizing commercial banking processes, revamping CX, and improving conversion rates by 60%.
  • Maintained internal tech docs on Azure DevOps Wiki; led sprint demos and scrum ceremonies; collaborated with stakeholders to shape requirements.
  • Drove a 40% reduction in Customer Effort Score via performance optimization; delivered via Azure CI/CD in an agile SDLC.

Software Engineer Intern — Capgemini

Feb 2021 – Apr 2021 · Java, SpringBoot, Angular
  • Led a team of 3 to build a sports‑equipment e‑commerce portal on Spring.
  • Owned Oracle 11g database ops and Angular front‑end features.
  • Implemented unit/integration tests with JUnit and Mockito.

Projects

Cross‑Modal Attention Fusion for Longitudinal Alzheimer’s Trajectory Modeling

Built a tri‑modal model integrating MRI (3D‑CNN), cognitive sequences (GRU), and genetics (MLP) with cross‑modal attention, achieving ROC‑AUC 0.9932 and precision 0.947. Ran ablations across 7 fusion variants; found genetic‑cognitive pairing most predictive and analyzed trade‑offs from modality‑specific noise and missingness.

PyTorch 3D CNN · GRU · MLP Cross‑modal attention

Transformers for Failure Prediction in HPC System Logs

Developed and evaluated LLM‑style models (minGPT, BERT) for anomaly detection in HPC logs, improving prediction accuracy by 15% on the System 20 dataset (LANL). Implemented probabilistic inference and diverse attention mechanisms.

BERT · minGPT Anomaly detection Probabilistic inference

Deep Learning Image Captioning for Accessibility

Designed a Merge‑Model architecture combining CNN and LSTM for image features and sequence generation. Compared 5 pre‑trained CNNs (VGG16/19, ResNet50, InceptionV3, Xception) and improved BLEU score.

CNN‑LSTM BLEU evaluation Accessibility

Skills

Languages

Java, Python, JavaScript, HTML, CSS, C++, SQL

Frameworks & Libraries

PyTorch, VueJS, React, SpringBoot, Keras, NumPy, Pandas

Databases

Oracle, MongoDB, MySQL, SQLite

GenAI Tools & LLMs

Hugging Face, GPT, BERT, LangChain, Prompt Engineering, RAG

DevOps & Tools

Azure (Microsoft Certified AZ‑900), Docker, Kubernetes, Git, Postman, Eclipse

Education

University of California, Davis — M.S. Computer Science

Sep 2023 – Present · GPA: 4.0

Relevant Courses: Machine Learning & Discovery, Advanced Deep Learning, Vision & Language Research

Savitribai Phule Pune University — B.E. Computer Engineering

Jul 2017 – Jul 2021 · GPA: 3.74

Relevant Courses: Engineering Mathematics, Discrete Mathematics, Data Structures & Algorithms

Contact

Open to software engineering roles (AI/ML, full‑stack) and research collaborations. Best way to reach me is email.

Email Tejas Connect on LinkedIn GitHub