AI Hospital — Multi-Agent Medical System
Full-stack virtual hospital with LangGraph agent orchestration, RAG-enhanced diagnostics, and persistent patient records in PostgreSQL.
PythonFastAPILangGraphPostgreSQL+3
Abhyuday SharmaBuilding Intelligent Systems.
Third-year Computer Science Engineering student at RV College of Engineering (CGPA: 8.81/10). Skilled in AI/ML, agentic systems, and software development through internships at Samsung PRISM, C-DOT, and Central Water Commission. Proficient in Python, LangGraph, DSA, and dynamic programming.
Abhyuday Sharma|Software Engineer|Bengaluru, India
Selected Work
Production-grade AI systems and agentic workflows
AI-Lawyer Legal Case Platform
Comprehensive legal case management system with ML classification, LangGraph agents, RAG retrieval, and Telegram bot integration.
PythonFastAPILangGraphReact+3
PostDisaster AI Agentic Alert System
Full-stack disaster response system analyzing satellite imagery with Gemini Vision and 5 specialized CrewAI agents.
PythonCrewAIFastAPIReact+1
HAR Deep Learning Comparison
Research study comparing CNN-LSTM, BiLSTM, Attention, and Transformer architectures for Human Activity Recognition on UCI-HAR dataset.
PythonPyTorchDeep LearningTransformers
Experience
Building real-world systems with measurable impact
Research Intern
Samsung PRISM
Jan 2026 – PresentBengaluru, India
- Developing Screen2LaTeX Conversion (SLC) system using lightweight neural networks to convert screenshots and PDFs into portable LaTeX code
- Engineering pipeline to detect and extract text, equations, tables, and figures using bounding box localization for automated document reconstruction
- Implementing CNN/CRNN architectures fused with OCR techniques to optimize document layout analysis and image-to-text generation
PythonCNNCRNNOCRNeural NetworksLaTeX
Agentic AI Engineer Intern
Centre for Development of Telematics (C-DOT)
- Architected an autonomous SOC analyst using LangGraph and Python to orchestrate Tier-1 security investigations by correlating multi-host telemetry
- Engineered a RAG pipeline with ChromaDB and OpenSearch; grounded LLM reasoning in MITRE ATT&CK (350+ techniques), achieving 100% recall on curated synthetic attack sets
- Deployed Dockerized microservices with Streamlit dashboard and hybrid Ollama/Gemini backend, reducing incident triage time by 97% in internal testing
PythonLangGraphChromaDBOpenSearchDockerStreamlitRAG
Data Science Intern
Central Water Commission (SWRMO Division)
- Engineered production-grade ML pipeline with lag/rolling/cyclic features integrated with IMD rainfall data, achieving 88.15% accuracy within ±0.15m and MAE = 0.0942m
- Benchmarked multiple LSTM architectures; validated XGBoost for similar accuracy at lower compute
- Maintained reproducible codebase with documented assumptions: github.com/AbhyDev/cwc-flood-forecasting-ml
PythonXGBoostLSTMPandasTime SeriesFeature Engineering
Project Intern
Bhatiyani Astute Intelligence
May 2025 – Jul 2025Bengaluru, India
- Built an offline NFC-based tap-to-pay wallet for visually impaired users using ESP32-S3, ESP-IDF, and AES encryption
- Integrated haptic PIN entry for 2FA, EEPROM-backed credential storage, and physical key support for secure screenless transactions
ESP32ESP-IDFC/C++NFCAES EncryptionEmbedded Systems
Education
RV College of Engineering
B.E. Computer Science & Engineering — CGPA: 8.81/10 (till date)
Bengaluru, Karnataka, IndiaOct 2023 – Aug 2027 (expected)
Delhi Public School
Higher Secondary Education
Bhopal, Madhya Pradesh, IndiaMay 2023
Research
Publications and technical writing
Benchmarking Neural Architectures for Smartphone-Based Activity Classification
Co-Author — BiLSTM-Attention achieved 93.38% accuracy with 12.5× fewer parameters than Transformer•2025Paper
AI Hospital: Multi-Agent Medical Consultation System
Co-Author — IMCA architecture reducing token consumption by 81%, polyglot persistence with 9-specialty RAG•2025Paper
FileForge: Cloud-Native Distributed File Processing Platform
Co-Author — Near-linear throughput scaling with sub-50ms API latency via FastAPI, RabbitMQ, MinIO•2025Paper
Testimonials
What collaborators say
"His performance was good and he was found to be sincere and hardworking."
S
Saurabh Basu
Scientist 'E' & Group Head, DSA-ESOC @ C-DOT (Centre for Development of Telematics)
"During the training his performance and behaviour was found to be excellent."
S
Sachin Gupta
Dy. Director @ Central Water Commission
Achievements & Certifications
Recognition and verified skills
2nd Place — Rewind and Recode Hackathon
IIIT Bhubaneswar | State: Karnataka
Secured 2nd position in Karnataka for presenting Multi-Agent AI Hospital Simulation
2025
CCNA: Introduction to Networks
Cisco
Networking fundamentals, IP addressing, and network infrastructure
Jun 2025
View CertificateAI Agents Bootcamp: LangChain, RAG & GPT
Udemy
Comprehensive training on building AI agents, RAG systems, and LLM applications
2025
View CertificateWeb Development Boot Camp
RVCE Coding Club
Full-stack web development fundamentals and modern frameworks
2024
View CertificateCS50: Introduction to Computer Science
Harvard University
Foundational computer science and programming principles
2024
Practical Deep Learning for Coders
fast.ai
Practical deep learning with PyTorch and modern architectures
2024
Stack
Technologies I work with daily
DSA (100+ LeetCode)OOPOSDBMSComputer NetworksMachine LearningDeep LearningAgentic WorkflowsRAGNLPComputer VisionPythonC/C++JavaScriptSQLHTML/CSSPyTorchLangGraphLangChainCrewAIscikit-learnXGBoostPandasNumPyTransformersCNNsLSTMsfast.aiFastAPI (REST, JWT)FlaskStreamlitPostgreSQLSQLAlchemyChromaDBRedisDockerGitGitHub ActionsAWS (EC2, S3)LinuxPostmanVS Code
Get in Touch
Let's build something together