AI Research Intern
DaRyte Ventures Pvt Ltd
Built role-based ReAttire workflows and backend/API features, and authored two AI-literacy and corporate-readiness guides for veteran reskilling.
One-page ATS résumé / updated 8 October 2026
Sanidhya Singh
AI/ML research engineering with a focus on medical AI and MedTech.
Profile
MedTech-focused AI/ML research engineer and UCD MSc candidate building measurable systems across clinical evidence retrieval, biomedical modelling and medical data analysis.
Experience
Research & Development Intern
IIT Delhi (A.I.C.)
Developed an AI + IoT crop-surveillance prototype using YOLOv11 and LoRaWAN, including a lighting-related model/data iteration.
Selected projects
IrishClinicalRAG
Clinical evidence retrieval · Python, FastAPI, BM25, BGE
Built a clinical-guidance RAG pipeline over 8 HSE/NCEC documents and 692 section-aware passages, combining a 692 × 384 semantic index, reciprocal-rank fusion, reranking and exact chunk citations. Recorded 29 passing tests, 75.36% statement coverage and approximately 2.0 s warm CPU retrieval.
VoxRelay Offline Multilingual Subtitles
Speech engineering · Python, PyTorch, torchaudio
Built an offline multilingual speech-to-English subtitle system integrating streaming ASR, translation and subtitle rendering. Engineered a 16 kHz, 512-point FFT, 80-filter log-Mel core that matched torchaudio within an approximately 1.16 × 10^-5 maximum feature difference.
APObind Binding-Site Experiments
Biomedical graph learning · ESM-2, PyTorch Geometric
Investigated residue-level ligand-binding prediction over 10,599 apo-holo protein pairs using ESM-2 embeddings and Graph Attention Networks, with evaluation structured around PR-AUC, ROC-AUC, F1, MCC and class imbalance.
Enhanced Compression Model
IEEE INSTCon 2026 · PyTorch, CompressAI
Co-developed a hyperprior neural image codec and ablation pipeline that reported 0.176 BPP, 30.882 dB PSNR and 0.985 MS-SSIM on the 24-image Kodak benchmark.
Education
MSc Artificial Intelligence for Medicine & Medical Research
University College Dublin
Postgraduate study in medical AI and biomedical research; ongoing.
B.Tech (Hons), CSE — IoT & Intelligent Systems
Manipal University Jaipur
Honours programme focused on computer science, IoT and intelligent systems.
Methods and tools
Medical AI
Clinical evidence retrieval, biomedical modelling, protein language models, graph learning and evidence-grounded generation.
Machine learning
Python, PyTorch, PyTorch Geometric, CompressAI, transformers, computer vision, GNNs, evaluation and ablation studies.
Engineering
FastAPI, BM25, dense retrieval, reranking, Next.js, Node.js, SQL, Git, Linux, Docker and LoRaWAN.
Leadership and certifications
SAP National Hackathon finalist. Selected completed learning includes Google Foundations of Data Science, IBM SQL for Querying Databases, IBM Linux Commands and Shell Scripting, and LearnQuest Foundations of Secure IoT Architecture.