Hire Teja R. - Generative AI Engineer

At Accelerant, I design and develop Accelerant AI Hub, a centralized enterprise Generative AI platform that brings together multiple LLMs, AI assistants, enterprise knowledge, AI agents, reusable skills, MCP integrations, scheduled workflows, and AI applications within a unified experience. I work with Azure OpenAI, Azure AI Foundry, Azure AI Search, LangGraph, LangChain, Python, and FastAPI to build scalable, grounded, and context-aware enterprise AI solutions. I design agentic RAG and multi-agent architectures using LangGraph and ReAct, enabling intent-aware routing across specialist agents, targeted RAG, enterprise tools, connectors, and MCP capabilities. These workflows support dynamic tool calling, iterative retrieval, parallel agent execution, response synthesis, conversational context, and agent memory for complex multi-step enterprise interactions. I also work on multi-LLM orchestration across GPT, Claude, Grok, and Gemini, enabling flexible model selection and routing for different enterprise use cases. I develop reusable Skills, enterprise Knowledge Base capabilities, MCP integrations, and scheduled AI workflows that allow users to combine models, knowledge, tools, and agents for specialized AI solutions. I focus on building secure, production-ready AI platforms using Microsoft Entra ID, OAuth 2.0, JWT, RBAC, delegated authorization, and permission-aware retrieval. I also contribute across FastAPI services, MuleSoft integrations, React, Docker, Azure Container Apps, CI/CD, observability, evaluation, and AI governance to deliver scalable and reliable enterprise AI capabilities.

At PennyMac, I design and implement Generative AI solutions that are transforming financial services and customer engagement. I develop and deploy GPT-4,5 powered chatbots with LangChain, LangGraph reducing customer service resolution times by 50% and improving cost efficiency. I also build retrieval-augmented generation (RAG) pipelines with ElasticSearch and Pinecone, which increase the accuracy of loan eligibility insights and financial data responses by 30%. These AI-driven systems directly support smarter decision-making across mortgage servicing, loan processing, and financial risk assessments. I work on developing and fine-tuning machine learning models in TensorFlow, PyTorch, and Scikit-learn for credit risk analysis, anomaly detection, and loan eligibility prediction, strengthening fraud prevention and reducing default risk. To ensure transparency and compliance, I apply SHAP and LIME for model interpretability and integrate models with FastAPI and Flask, delivering scalable APIs within PennyMac’s financial platform. In addition, I automate ETL pipelines for real-time data ingestion with Spark and Kafka, reducing reporting latency from hours to minutes. I also create executive dashboards in Tableau and Python (Matplotlib, Plotly) for real-time monitoring of fraud, credit risk, and AI model performance. By collaborating with finance, compliance, and risk analytics teams, I ensure that AI-driven insights meet regulatory requirements and business goals, driving measurable improvements in efficiency, accuracy, and customer experience.

At Elevance Health, I engineered AI/ML solutions for healthcare analytics, developing and fine-tuning supervised, unsupervised, and deep learning models (CNNs, RNNs, Transformers) to support medical imaging, patient outcome forecasting, and NLP on large-scale clinical datasets. By building robust end-to-end ML pipelines with MLflow, Airflow, and Spark, I streamlined model development and deployment while ensuring reproducibility, scalability, and strict HIPAA compliance. I led the implementation of LLM-powered Retrieval-Augmented Generation (RAG) pipelines using Hugging Face Transformers, FAISS, and Pinecone, enabling advanced healthcare question-answering and clinical document summarization. I also built NLP pipelines (spaCy, NLTK, Transformers) for sentiment analysis, entity recognition, and patient record insights, improving clinical decision support. To maximize performance, I optimized models with quantization, pruning, and GPU acceleration, reducing inference latency and boosting reliability in production environments. Collaborating closely with clinicians, product teams, and DevOps engineers, I integrated AI models into enterprise healthcare platforms, enhancing predictive analytics, patient care insights, and real-time monitoring. My role also included establishing CI/CD pipelines (Jenkins, GitHub Actions, Terraform) for continuous retraining, testing, and deployment, ensuring reliable model governance. Beyond delivery, I mentored junior engineers and contributed to research on transfer learning and federated learning, driving Elevance Health’s innovation in AI-driven healthcare transformation.