Gandham Gnanendra

• Designed and implemented robust pipeline for attribute-value extraction from enterprise documents using LLMs.
  Developed modular ChatJTI class in Python to support token handling, document upload, and dynamic prompt injection.
  Executed hybrid retrieval (BM25 + Vector + RRF) for accurate enterprise document search.
  Enhanced production reliability through error handling, retry logic, and logging mechanisms.
  Deployed solution in Azure Databricks with Azure Key Vault integration for credential management.
  Created internal KB Chatbot utilizing FastAPI, Streamlit, ChromaDB, and OpenAI embeddings.
  Enabled natural language queries for HR and company policy documentation.
  Implemented document ingestion and embedding pipelines within Azure Container Apps.

• Designed and implemented tool for parsing and analyzing log files.
  Integrated Hugging Face models for natural language processing and summarization.
  Developed custom GPT-4 model to predict real-time log failures.
  Automated log parsing process to efficiently detect issues.
  Conducted performance testing and optimization for real-time application.
  Collaborated with stakeholders to define key use cases and enhance functionality.

• Developed aspect extraction model to analyze consumer reviews effectively.
  Utilized sentence transformers to parse multiple sentences from reviews.
  Extracted noun chunks to derive contextual meanings from sentences.
  Mapped extracted context to predefined dictionary keywords to identify final aspects.
  Designed and managed models using Python and transformer architecture.
  Created sentiment analysis model to evaluate customer feedback for product improvement.
  Implemented sentiment analysis techniques to classify feedback as positive, negative, or neutral.
  Leveraged technologies such as NLTK, spacy, and Google Cloud for project execution.

Designed Root Cause Analysis system to predict network failures effectively.

Implemented algorithms for tracing and anticipating network outages.

Conducted root cause analysis integrating data from diverse network sources.

Developed reporting tools delivering insights on predicted failures.

Collaborated with network engineers to validate and enhance system outputs.

Executed algorithms assessing pricing trends and competitive landscape.

Forecasted Buy Box and Sales prices through advanced predictive modeling.

Advised clients on pricing strategies derived from predictive analytics.