Bonkur Harshith Reddy
Applied AI Research | Cloud Engineering | DevOps | MLOps
Hyderabad
harshithreddy.dev@gmail.com
Computer Science graduate passionate about AI/ML research, cloud computing, and DevOps. Co-author of 3 IEEE research publications, with interests spanning deep learning, computer vision, and applied AI. Beyond research, I'm interested in cloud-native technologies, infrastructure automation, CI/CD, and MLOps, with a focus on bridging the gap between machine learning research and scalable, production-ready systems. Always curious about emerging technologies, open-source projects, and opportunities to collaborate with people building impactful technology.
AI Intern
CSRBOX (IBM SkillsBuild)
- Developed Python and Pandas ETL pipelines to process, transform, and validate 10,000+ disaster-aid requests, improving data accuracy by 30%. Integrated REST APIs and automated validation workflows supporting 100+ concurrent users.
Intern
IITM Pravartak Technologies Foundation
- Developed multilingual Generative AI and RAG components using Python, FastAPI, LangChain, and FAISS for retrieval across 1,000+ agricultural documents. Optimized semantic search, retrieval, and caching pipelines to reduce query latency to under 2 seconds for 500+ daily users.
Founding Engineer
Cropyaan Pvt Ltd
- Develop Python backend services, REST APIs, and AWS-integrated application workflows. Work across development, testing, debugging, and deployment, with a focus on CI/CD automation, containerized services, reliability, and maintainable software.
Planner–Executor Architecture for Teledermatology Triage with Safety Guardrails
Co-authored an IEEE research paper proposing DermaSightX, a safety-focused AI architecture for teledermatology triage. Separates vision and language model perception from clinical decision-making through schema-validated observations, deterministic safety guardrails, mandatory human oversight, and auditability. The paper reports 89.6% classification accuracy and 100% recall for emergency cases.
Automated Parkinson’s Disease Identification and Diagnosis via an Advanced Deep Learning Framework
Co-authored an IEEE research paper investigating MRI-based Parkinson’s disease identification using a multi-stage deep learning framework. Combines U-Net brain segmentation, denoising autoencoders, EfficientNet feature extraction, and spatial attention. The proposed framework reports 99.4% classification accuracy on the study dataset.
Lightweight Model for Early and Multi-Class Detection Using CNN for Alzheimer’s Disease
Co-authored an IEEE research paper exploring a lightweight convolutional neural network for four-class Alzheimer’s disease severity classification using brain MRI scans. The model uses 277,956 trainable parameters and was evaluated on 7,536 MRI scans. The paper reports 99% overall classification accuracy.
Bachelors of Technology in Computer Science and Engineering
Vellore Institute of Technology, Andhra Pradesh
AWS Certified Solutions Architect – Associate
Amazon Web Services (AWS)
Sept 2026Redis
Sept 2026Azure AI Apps and Agents Developer Associate
Microsoft
Aug 2026Microsoft
Jul 2026Microsoft
Jul 2026MongoDB Certified Developer, Associate (C100DEV)
MongoDB
Oct 2025MongoDB Certified DBA Associate (C100DBA)
MongoDB
Sept 2025AWS Certified Cloud Practitioner
Amazon Web Services (AWS)
Jun 2025NPTEL
Nov 2025NPTEL
Mar 2023NPTEL
Aug 2022
Attendee · Oct 24, 2026