Senior AI Engineer (LLMs, RAG, Agents)

Kevin Weithers

Senior AI engineer Generative AI Agentic AI LLMs RAG GraphRAG Full-Stack Engineering Santa Clara, CA ***************@*****.*** 213-***-**** LinkedIn PROFESSIONAL SUMMARY

Senior AI Engineer with 8+ years of experience designing, building, and deploying production-grade AI, machine learning, voice AI, and cloud-native software platforms across enterprise environments. Proven expertise in Large Language Models (LLMs), Agentic AI, Multi-Agent Systems, Retrieval-Augmented Generation (RAG), GraphRAG, Conversational AI, Voice AI, NLP, Machine Learning, Data Engineering, and Full-Stack Development.

Experienced leading the end-to-end lifecycle of AI products, from data architecture and model development through production deployment, observability, governance, and continuous optimization. Strong track record of transforming fragmented enterprise knowledge into intelligent systems that automate research, streamline workflows, improve operational efficiency, and enhance user experiences through both text and voice-based interactions.

Hands-on experience building cloud-native AI platforms on AWS using Kubernetes, Terraform, Docker, Kafka, Airflow, and modern observability stacks while integrating enterprise systems such as Salesforce, HubSpot, Intercom, Stripe, and Twilio into intelligent automation and conversational AI workflows. Unique career progression from Data Scientist to Data Engineer to Machine Learning Engineer to Senior AI Engineer, providing deep expertise across the entire modern AI stack, including data platforms, machine learning systems, generative AI applications, infrastructure, and customer-facing products. CORE SKILLS

• AI/ML: Generative AI, Agentic AI, LLMs, RAG, GraphRAG, Voice AI, NLP, LangChain, LangGraph, LlamaIndex, OpenAI, Claude, ElevenLabs

• Backend: Python, FastAPI, Node.js, REST APIs, Microservices, WebSockets

• Frontend: React, Next.js, TypeScript

• Cloud & DevOps: AWS (EC2, S3, Lambda, RDS, EKS, SQS), Docker, Kubernetes, Terraform, GitHub Actions, Prometheus, Grafana, OpenTelemetry

• MLOps & Production AI: Model Versioning, Experiment Tracking, Model Registry, CI/CD for ML, Model Monitoring, Drift Detection, A/B Testing, Evaluation Frameworks

• Data: PostgreSQL, MySQL, MongoDB, Redis, Kafka, Airflow, Spark, ETL/ELT

• Integrations: Twilio, Salesforce, HubSpot, Intercom, Stripe PROFESSIONAL EXPERIENCE

Cognizant – Santa Clara, CA

Senior AI Engineer Jun 2024 – Present

• Lead architecture and development of enterprise Generative AI and Agentic AI solutions focused on intelligent search, knowledge management, workflow automation, decision support, and conversational experiences. Partner directly with stakeholders to identify high-impact use cases where AI can improve operational efficiency and reduce manual effort across business functions.

• Architect production-grade Retrieval-Augmented Generation (RAG) platforms that enable employees to access enterprise knowledge through natural language interactions. Designed complete retrieval pipelines incorporating document ingestion, semantic indexing, embedding generation, vector search, reranking, prompt orchestration, and response generation using modern LLM frameworks and cloud-native services.

• Established end-to-end MLOps lifecycle on Databricks and AWS environments, including experiment tracking, model versioning, automated retraining pipelines, and production deployment governance using CI/CD and model registries.

• Implemented production model monitoring systems tracking data drift, performance degradation, and inference latency, with automated alerting and rollback strategies to ensure model reliability in production environments.

• Built and deployed LLM and RAG pipelines on Databricks and AWS-based distributed data environments to support scalable AI workloads.

• Designed and implemented GraphRAG architectures to solve complex enterprise information retrieval challenges involving highly interconnected business entities, documents, and operational processes. Combined vector retrieval with knowledge graph traversal to improve contextual understanding and provide more explainable responses for multi-hop business questions.

• Developed multi-agent AI systems using LangGraph, LangChain, OpenAI models, and Claude where specialized agents perform planning, retrieval, reasoning, validation, and response synthesis. Enabled AI applications to handle complex multi-step workflows that traditionally required manual coordination across teams and systems.

• Built scalable document ingestion and processing frameworks capable of handling large collections of structured and unstructured enterprise content including technical documentation, operational procedures, contracts, policy documents, SharePoint repositories, and knowledge base articles. Implemented metadata extraction, semantic chunking, indexing, and retrieval optimization strategies to maximize downstream answer quality.

• Addressed common enterprise AI challenges including hallucinations, inconsistent outputs, incomplete retrieval, and lack of source attribution by implementing retrieval validation, response grounding, citation generation, confidence scoring, and evaluation frameworks that improved trustworthiness and adoption of AI-powered solutions.

• Designed and implemented Voice AI assistants that combined enterprise knowledge retrieval with real-time conversational capabilities, enabling users to interact with AI systems through natural voice conversations instead of traditional text interfaces.

• Built end-to-end voice agent architectures integrating Twilio Programmable Voice, OpenAI models, ElevenLabs text-to-speech services, and retrieval pipelines to support customer service, internal helpdesk, and knowledge-assistant use cases.

• Developed real-time conversational workflows involving speech-to-text processing, intent understanding, retrieval augmentation, reasoning, and natural voice response generation, creating seamless voice experiences while maintaining conversational context.

• Implemented low-latency streaming architectures using WebSockets and event-driven communication patterns to support real-time voice interactions and improve responsiveness across conversational AI applications.

• Integrated enterprise RAG systems with Voice AI platforms, allowing users to access internal documentation, operational procedures, and organizational knowledge through phone-based and web-based conversational interfaces.

• Designed cloud-native AI services on AWS leveraging EKS, Lambda, S3, RDS, and SQS to support scalable deployment of LLM applications, retrieval systems, voice assistants, and agentic workflows.

• Automated infrastructure provisioning using Terraform, enabling repeatable deployment of AI environments across development, testing, and production while improving operational consistency and reducing manual infrastructure management.

• Containerized AI applications using Docker and Kubernetes, allowing engineering teams to deploy, scale, and maintain LLM-powered services efficiently under varying enterprise workloads.

• Established observability frameworks using Prometheus, Grafana, OpenTelemetry, and ELK Stack to monitor application performance, retrieval latency, infrastructure health, model behavior, and API reliability.

• Implemented CI/CD pipelines through GitHub Actions to automate testing, validation, container builds, infrastructure deployment, and release management processes for AI applications.

• Integrated AI platforms with Salesforce, HubSpot, Intercom, and Twilio services, enabling conversational agents to access customer context, support histories, operational data, and communication channels within enterprise workflows.

• Led model assessment initiatives across GPT, Claude, and open-source foundation models, evaluating tradeoffs involving reasoning quality, latency, operational cost, scalability, and deployment constraints to support architecture decisions.

• Established evaluation methodologies to assess retrieval quality, response relevance, factual consistency, and user satisfaction. Collaborated with stakeholders to define measurable acceptance criteria and continuously improve AI performance through iterative testing.

• Mentored engineers on LLM application development, agentic architectures, prompt engineering, retrieval systems, AI observability, and production deployment best practices. Amazon – Chicago, Illinois

Senior Machine Learning Engineer October 2021 – June 2024

• Designed and deployed enterprise Voice AI solutions leveraging Amazon Connect, Amazon Lex, Twilio, and ElevenLabs to automate customer support and internal service workflows. Integrated LLM-powered conversational agents with retrieval-based knowledge systems, enabling users to access operational information through natural voice interactions while maintaining context across multi-turn conversations.

• Built AI-powered knowledge assistants using Amazon Q and Retrieval-Augmented Generation

(RAG) architectures to help engineering and operations teams quickly discover technical documentation, troubleshooting guides, and organizational knowledge. Integrated Amazon Q with internal repositories and AWS services, reducing time spent searching across distributed documentation and improving knowledge accessibility across teams.

• Designed and deployed machine learning solutions supporting customer operations, business intelligence, process optimization, and strategic decision-making initiatives across multiple enterprise functions.

• Partnered with business leaders, product owners, and engineering teams to identify operational challenges that could be addressed through predictive analytics and machine learning, translating business objectives into scalable technical solutions.

• Developed predictive models leveraging customer behavior data, transactional information, and operational records to support forecasting, prioritization, segmentation, and optimization initiatives.

• Built large-scale feature engineering pipelines using Python, SQL, Spark, PostgreSQL, Kafka, and cloud-native data platforms, enabling consistent generation of high-quality model inputs from diverse enterprise datasets.

• Developed reusable machine learning workflows that standardized model training, validation, deployment, monitoring, and retraining processes, improving collaboration between data science and engineering teams.

• Created NLP-based analytical solutions for processing customer interactions, support documentation, and operational records. Applied modern language-processing techniques to identify recurring issues, emerging trends, and opportunities for service improvement.

• Designed automated ETL and ELT workflows using Airflow and cloud-based infrastructure to support analytical reporting and machine learning initiatives across multiple business domains.

• Worked extensively with large-scale structured and semi-structured datasets, addressing challenges related to data quality, data lineage, missing values, evolving business definitions, and source-system inconsistencies.

• Collaborated with software engineering teams to productionize machine learning capabilities through APIs, containerized services, and scalable deployment architectures.

• Established model monitoring and governance practices to detect performance degradation, feature drift, and data anomalies before they impacted downstream business processes.

• Implemented observability solutions using ELK, Prometheus, and Grafana to improve visibility into model performance, deployment health, pipeline reliability, and operational metrics.

• Participated in MLOps initiatives involving Docker, Kubernetes, CI/CD pipelines, automated testing, version control, and model lifecycle management.

• Evaluated emerging Generative AI technologies and LLM capabilities, conducting proof-of- concept initiatives that explored future opportunities for enterprise AI adoption.

• Mentored junior engineers and data scientists on machine learning engineering, production deployment strategies, stakeholder communication, and technical best practices. Salesforce – Chicago, Illinois

Data Engineer April 2018 – October 2021

• Designed and maintained enterprise-scale data platforms supporting analytics, reporting, machine learning, and operational decision-making across multiple business domains.

• Developed large-scale data ingestion and transformation pipelines using Kafka, Airflow, PostgreSQL, MongoDB, MySQL, Spark, and cloud-native technologies to support growing analytical workloads.

• Built scalable ETL and ELT frameworks that consolidated information from Salesforce products, customer platforms, operational systems, third-party integrations, and cloud storage environments.

• Partnered closely with data scientists, analysts, and business intelligence teams to provide reliable, production-ready datasets optimized for reporting, experimentation, and predictive modeling.

• Solved complex data quality challenges through implementation of validation frameworks, anomaly detection processes, reconciliation checks, and monitoring systems that improved trust in enterprise data assets.

• Designed data warehouse structures and analytical models capable of supporting both operational reporting requirements and advanced analytical workloads while maintaining scalability and performance.

• Led modernization initiatives that migrated legacy data processing workflows to cloud-native architectures, improving maintainability, operational efficiency, and scalability.

• Optimized large-scale Spark workloads through partitioning strategies, query tuning, indexing improvements, and storage optimization techniques that enhanced overall processing performance.

• Built monitoring and alerting systems providing visibility into pipeline health, latency trends, processing failures, and data quality issues, reducing operational risk and improving incident response.

• Implemented workflow orchestration solutions using Apache Airflow, improving reliability, scheduling flexibility, and maintainability of mission-critical data pipelines.

• Developed Looker dashboards and reporting solutions that enabled stakeholders to gain visibility into operational performance, data quality metrics, and business KPIs.

• Leveraged Redis caching strategies and optimized database architectures to improve performance of high-volume data services and reduce latency for frequently accessed datasets.

• Integrated enterprise data platforms with external business systems and SaaS applications, creating unified reporting and analytics capabilities across organizational functions. Enova International – Chicago, Illinois

Data Scientist May 2016 – April 2018

• Developed predictive analytics and machine learning solutions supporting customer analytics, portfolio monitoring, operational decision-making, and business growth initiatives.

• Applied statistical modeling, machine learning algorithms, and exploratory data analysis techniques to identify behavioral patterns, operational trends, and business opportunities hidden within large datasets.

• Built classification, regression, forecasting, and segmentation models designed to improve understanding of customer behavior and support data-driven decision making.

• Conducted extensive feature engineering and data preparation activities, transforming raw operational and customer data into actionable signals that improved model performance and analytical effectiveness.

• Partnered with business stakeholders, product teams, and operational leaders to understand analytical requirements and translate them into practical machine learning solutions.

• Designed experiments and analytical frameworks to evaluate business initiatives, measure outcomes, and support evidence-based strategic planning.

• Applied NLP techniques to extract insights from customer communications, support interactions, and unstructured operational documentation.

• Created dashboards, visualizations, and executive reporting solutions that translated complex analytical findings into actionable recommendations for technical and non-technical audiences.

• Improved model validation and testing practices through rigorous performance assessment methodologies, cross-validation techniques, and continuous monitoring approaches.

• Collaborated with engineering teams to operationalize analytical solutions and transition research models into production-ready systems.

• Promoted data-driven decision making throughout the organization by advocating analytical best practices and helping teams leverage data more effectively. EDUCATION

University of Chicago

Master of Science (M.S), Computer Science 2020 – 2021 University of Notre Dame

Bachelor of Science (B.S), Applied Mathematics and Statistics 2012 – 2016

Contact this candidate