Shreerama D.

About Me

Shreerama is an AI/ML Engineer experienced in LLMs, RAG pipelines, and Generative AI. Skilled in NLP, image processing, and end-to-end AI deployment, he builds real-time AI solutions using Python, PyTorch, and TensorFlow. Shreerama has published an IEEE research paper and contributed as a peer reviewer for NetACT 2025.

AI, ML & LLM

BERT Deep Learning GPT Generative AI LLM LangChain OpenAI API Pytorch Streamlit AI/ML

Frontend

Backend

Database

Workflow

Other

Autoencoders C CNN Classification Clustering Computer Vision Data Visualization Deployment EDA GAN Hyperparameter Tuning Image Classification Image Processing Jupyter LSTM LoRa NLP Nomic Embed Text Numpy Pandas Problem Solving Pycharm RAG RNN Replit SpeechRecognition Supervised Learning Tensorflow Unsupervised Learning

Work history

BulkBeings
BulkBeings
AI Engineer (Internship)
2025 - 2025
Chennai, India
  • Built a real-time NOTAM dashboard using LLMs, RAG, and Qwen model for aviation alerts.

  • Integrated multi-source data pipelines and implemented citation-backed responses.

  • Optimized prompt logic and deployed scalable RESTful APIs.

  • Built and deployed AI-powered solutions using Generative AI, LLMs, and RAG pipelines.

  • Developed intelligent systems for real-world applications, optimizing model performance and integrating custom AI models into scalable back-end architectures.

  • Deployed advanced Machine Learning solutions in a startup environment.

LLMs RAG Artificial IntelligenceData pipelinesRESTFul APIs Generative AI Performance OptimizationAI Model Intergration Machine LearningPython
Hindustan Aeronautics Limited (HAL)
Hindustan Aeronautics Limited (HAL)
AI Research Intern
2024 - 2024
Chennai, India
  • Improved the HiFiC model for high-quality, lossless image compression.

  • Enhanced Deep Learning performance through model tuning and error resolution.

  • Developed a GAN model aimed at enhancing image compression while preserving high visual quality.

  • Conducted data preprocessing, including resizing and normalization, to prepare a diverse dataset of images for model training.

  • Evaluated model performance using key metrics such as Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index (SSIM), ensuring effective compression without significant quality loss.

  • Collaborated with a multidisciplinary team to troubleshoot model issues, optimizing performance and gaining valuable insights into real-world AI applications.

Showcase

Cysinfo AI
Cysinfo AI

Fine-tuned LLaMA 3.1 using LoRA and domain-specific datasets to achieve high contextual accuracy for cybersecurity and ethical hacking queries, improving response accuracy for Kali Linux commands by 20%. Developed a web interface, integrating advanced NLP APIs, deployed the model on Ollama server, optimizing query processing latency by 30%. Tech stack: LLMs, Generative AI, AI development.

DocuSense
DocuSense

Designed an AI-powered documentation assistant using advanced RAG with LangChain, LLaMA 2, and Nomic Embed Text to enable real-time query resolution and contextual awareness. Developed and deployed an interactive application using Streamlit, integrated with the Ollama server. Tech stack: RAG, LLaMA 2, LangChain, Generative AI, AI development, Streamlit, NLP.

Voice Genius
Voice Genius

Developed an AI-driven voice assistant using OpenAI API and Speech Recognition. The application processes spoken questions and provides answers in both voice and text formats. Tech stack: NLP, Speech Recognition, API integration, deployment.

Beyond Skin Deep
Beyond Skin Deep

Implemented and optimized Deep Learning models for melanoma detection, focusing on hyperparameter tuning and performance evaluation. Tech stack: Deep Learning, Image Classification, Hyperparameter Tuning.

Education

Machine Learning
Machine Learning
Coursera
2023 - 2023
B.Tech Artificial Intelligence
B.Tech Artificial Intelligence
Dayananda Sagar University - India
2022 - 2025 (3 years)