Data Scientist & Full-Stack AI Developer

Data H Scientist LOKESH Full-Stack REDDY AI Developer

***************@*****.*** • +91-706******* • Rourkela, Odisha • github.com/LokeshAI669 • linkedin.com/in/hlokesh-reddy-6b4510368

PROFESSIONAL SUMMARY

CSE graduate (CGPA 8.02) with deployed AI systems in NLP (BERT), Computer Vision (YOLOv8, 94% confidence, 28+ FPS), and Generative AI (LLaMA 3, sub-2s API response). Achieved 88% ML accuracy and 30% inference speedup during internship. Proficient in Python, FastAPI, Next.js, and cloud deployment (AWS, Vercel, Render). EDUCATION

B.Tech — Computer Science & Engineering, Kalam Institute of Technology (BPUT) 2021 – 2025 CGPA: 8.02 Relevant coursework: Machine Learning, Data Structures, Database Systems, Statistics, Computer Vision WORK EXPERIENCE

Data Science Intern — Cognifyz Technologies Nov – Dec 2024

• Built wine quality classifier (Random Forest & XGBoost) on 1,599-sample UCI dataset; achieved 88% accuracy vs. 74% logistic regression baseline — a +14% improvement.

• Cut model inference time by 30% via feature selection (removed 4 low-importance variables); packaged with MLflow for reproducibility. Identified alcohol content & volatile acidity as top predictors. PROJECTS

Medical AI Assistant Next.js, FastAPI, Supabase, LLaMA 3, PostgreSQL, Vercel, Render

• Built full-stack medical app with JWT authentication, Retrieval-Augmented Generation (RAG)-powered symptom analysis, and AI chatbot; achieved sub-2s API response with rate limiting.

• Deployed frontend on Vercel and backend on Render with PostgreSQL; implemented role-based access control, secure session management, and REST API integration.

• Generated automated PDF medical reports and a fully responsive UI for mobile and desktop users. PAN Card Detection System YOLOv8, OpenCV Dataset: 500+ custom-labelled images

• Developed real-time document detector achieving 94% confidence at 28+ FPS on live webcam with bounding box and confidence score overlay.

• Fine-tuned YOLOv8 via transfer learning on 500+ custom-labelled images, reducing false positives by 40% compared to the base model.

• Built auto-crop-and-save pipeline to extract detected PAN card regions; added FPS monitoring and frame-level logging for performance benchmarking.

AI Resume Screening System BERT, spaCy, Scikit-learn, Streamlit

• Built NLP pipeline using BERT embeddings and spaCy for automated resume parsing, skill extraction, and job description matching via cosine similarity scoring.

• Reduced manual screening time by ~70% through automated candidate ranking; supports batch processing of multiple resumes against a single job description.

• Designed Streamlit recruiter dashboard to upload JDs/resumes, view ranked candidates with match scores, highlight skill gaps, and export shortlists as CSV.

TECHNICAL SKILLS

ML / AI: Scikit-learn, XGBoost, TensorFlow, BERT, YOLOv8, LangChain, RAG, LLMs, Prompt Engineering, CNN, RNN, LSTM, ARIMA

Languages & Data: Python (Advanced), SQL, NumPy, Pandas, Matplotlib, Seaborn, EDA, Feature Engineering, A/B Testing Backend & Deploy: FastAPI, Flask, Streamlit, Docker, MLflow, REST APIs, AWS, Vercel, Render, Supabase, PostgreSQL, Git, GitHub

CERTIFICATIONS

• Full Stack Data Science & AI — Naresh IT (2025)

• Data Science Intern Certificate — Cognifyz Technologies (Dec 2024)

• Data Analytics Job Simulation — Deloitte via Forage (2024)

• Fundamentals of Machine Learning and Artificial Intelligence — AWS Training & Certification (2026)

Contact this candidate