Career Path

Professional Experience

Principal Data & AI Engineer — GenAI Platform Lead

AT&T — Dallas, TX
January 2024 – Present
  • Own uptime and availability of AT&T's enterprise AI Gateway across production (SLA-bound with call-to-work ticketing and outage ownership) and high-visibility non-production environments — covering open-source model deployments, embedding models, classifiers, and connectivity to paid model providers (Azure OpenAI, OpenAI)
  • Lead software engineer for CDO Inferencing Services — scalable inference endpoints for LLMs, embedding models, and classifiers leveraging KServe, NVIDIA Triton, NVIDIA NIM, vLLM engine, and LiteLLM Proxy on Kubernetes (AKS and on-premises clusters with InfiniBand/OFED)
  • Contributed to the open-source LiteLLM project — shipped production fixes to the Presidio guardrail integration (eliminating redundant/unwanted analyzer calls) and extended Azure Content Safety and Azure Prompt Shield guardrail integrations to correctly handle model call payloads exceeding 10K characters
  • Lead software engineer for CDO Speech Services — unified speech and translation API and Python client library built on NVIDIA Riva server on Kubernetes
  • Deployed Kubeflow on AKS and led the strategic vision for integrating Generative AI into enterprise-level products; recognized as the official adopter of both KServe and Kubeflow at AT&T
  • Led a team of software engineers, data scientists, and ML engineers to develop and maintain scalable AI platforms; main code reviewer for all CDO Speech and Inferencing Services code (Python, FastAPI, Docker, Helm, Kubernetes manifests, Azure deployment pipelines)
  • Lead software engineer for internal AutoGen Studio deployment — implemented storage persistence via Azure PostgreSQL, authentication/authorization via Azure App Registration, and a data governance layer to segregate group user data
  • Mentored junior engineers on CI/CD with Azure DevOps, Azure Web App Services for containers, RAG platform usage, and Kubernetes best practices

Senior Big Data Software Engineer

AT&T — Dallas, TX
November 2020 – December 2023
  • Lead software engineer for a configurable Retrieval-Augmented Generation (RAG) platform to deploy question-and-answer services on domain-specific documents, leveraging LangChain, OpenAI, open-source LLMs, fine-tuned embedding models, NVIDIA Triton, Milvus vector store, CosmosDB, Azure Cognitive Search, and Kubernetes — ensuring compliance with data privacy and security standards
  • Developed a Databricks job monitoring agent to supervise job status and projected cost based on job duration and compute specifications
  • Led design, development, and deployment of a data insights platform using AKS and Databricks to support product/pricing strategies; supervised Java developers and data engineers; developed and optimized batch and real-time data pipelines using Spark as consumer and producer for Kafka/EventHub
  • Designed and implemented ETL processes for data cleaning and standardization on Azure Databricks for an in-house product strategy platform
  • Led the platform migration from AWS to Azure, including Azure Repos, CI/CD Pipelines, and a custom resource management infrastructure system using the Java Azure SDK
  • Lead engineer for WatchtowerAI, an internal AT&T event-based monitoring platform — responsible for new feature development, enhancements, and code validation

Senior Software Engineer

AT&T — Dallas, TX
December 2019 – November 2020
  • Developed full-stack solutions (ReactJS, MariaDB, MongoDB) supporting automation for Network Engineering & Operations (NEO), streamlining equipment configuration creation
  • Implemented an ETL process for consolidating equipment inventory databases using Python and relational database connectors
  • Created custom Docker images and migrated POC applications to Docker; shared knowledge and best practices with the team

Data Scientist / Engineer

AT&T — Plano, TX
January 2019 – December 2019
  • Developed algorithms for HBase data processing using Java; created custom data parsers using Flink, Spark, and MapReduce via Java APIs
  • Developed a Python Folium/Leaflet visualization tool for algorithm evaluation results presented to leadership; conducted data exploration using Pandas on Vertica and Teradata
  • Implemented multiprocessing in Python for correlation validation and efficient large-scale data analysis

Software Engineer

AT&T — Dallas, TX
July 2017 – December 2018
  • Developed and tested a route optimization platform for repair/install technicians using OR-Tools (Python), saving $220 million annually; implemented street-level routing and job-specific prioritization via the Grasshopper API
  • Implemented ETL processes to support technician sharing between adjacent geographies; contributed to bash-based server setup automation

Student Co-Op Technical I

AT&T — El Paso, TX
March 2016 – January 2017
  • Automated repetitive tasks improving efficiency by 75% using VBA and Python; managed and validated data for 5,000 employees supporting a notification platform migration
  • Designed organization-specific curriculum, HTML/CSS access portals, and SQL/PHP databases; monitored and classified cellular network sites by performance metrics