Kush Batra
Director, Generative AI & Engineering
AI & Engineering · Platform & Cloud · DevOps/SRE — 16+ years turning emerging technology into production-grade engineering impact.
Strategy at the top, hands dirty underneath.
Engineering and technology leader with 16+ years across Generative AI, platform engineering, cloud infrastructure, SRE/DevOps and enterprise data platforms, in large-scale and regulated organizations.
Most recently Director, Generative AI & Engineering at Capital One, leading Engineering, DevOps and MLE teams on enterprise AI programs — taking AI from experiment to secure, governed, production-ready capability with reusable architectures, evaluation, observability and adoption models.
Combines strategic leadership with hands-on depth in AI/ML, RAG, cloud architecture, Kubernetes, distributed systems and DevOps/SRE. Delivered enterprise AI capabilities adopted by 30,000+ users, a code-review platform spanning 50,000+ repositories, and $1.8M+ monthly infrastructure savings.
I partner with CTOs and senior technology leaders, manage managers, hire and mentor strong engineers, and stay hands-on in architecture, design and technology choices.
Technology Leadership
AI & Engineering
Cloud & Platforms
Reliability & Operations
Application & Data
Sixteen years, one throughline: ship it, scale it, keep it reliable.
Director, Generative AI & Engineering
Aug 2026 — Oct 2026- Led Engineering, DevOps and MLE teams across three enterprise AI programs, partnering with senior technology leadership, Risk, Cyber and the Model Risk Office on priorities, governance and adoption.
- Defined engineering strategy for AI-powered developer experience and enterprise GenAI capabilities, shaping roadmaps with global ML and engineering leadership.
- Contributed to the CEO-level annual "How We Build Software" strategy exercise, identifying GenAI and cyber solutioning gaps and shaping the next year's initiative scope.
- Set direction for AI-assisted SDLC capabilities — code and PR review, headless test-case generation, technical documentation and engineering workflows — with evaluation, observability and security controls.
- Aligned team size, skill mix and hiring plans with the upcoming year's delivery demand, working with internal and external stakeholders.
Senior Manager, GenAI Initiatives
Aug 2023 — Aug 2026- Scaled AI developer tools — Windsurf, Claude Code, Cowork and Codex — to 30,000+ users, with access controls, identity integration, token limits, sensitive-data guardrails and OpenTelemetry productivity telemetry.
- Built and scaled an agentic code-review platform from scratch to 50,000+ repositories and 350K+ pull requests per month, at under $4 median cost per PR.
- Built a GenAI developer assistant chatbot adopted by 15,000+ engineers.
- Led development of DeX, a multi-tenant agentic RAG chatbot grounded in product documentation and historical Slack conversations; scaled to 70+ platforms, serving 90% of queries and resolving 80% of repeated queries.
- Architected a production RAG knowledge platform on AWS EKS with OpenSearch, semantic retrieval, reranking and LLMs.
- Secured enterprise LLM access by integrating Claude Code with AWS Bedrock via API Gateway and OAuth, and led the evaluation and launch of Windsurf with cybersecurity, risk and compliance.
- Led headless AI initiatives for test-case generation, code review, documentation refresh and change risk-impact analysis.
Senior Engineering Manager, Central Data
2021 — 2023- Led SRE, DevOps and platform engineering for central data platforms totalling 55PB, including a 15PB data warehouse, and owned business-unit vision, P&L, cost and resource planning.
- Delivered $1.8M+ monthly savings through cloud modernization, compute migration, data-transfer optimization and automated cost controls.
- Modernized the data platform from Hadoop/Hive on EC2 to Kubernetes-based Presto on EKS, self-managed Kafka to managed Kafka, and long-running EMR workloads to transient/serverless EMR.
- Automated JIRA-driven provisioning (days to about one hour) and access management (2–7 days to about 10 minutes); cut patch-cycle turnaround from about three months to one week.
- Built Kubeflow MLOps pipelines and supported GPU-backed OCR workloads; introduced Databricks and managed data-lake access and governance.
Specialist
2021- Led development, DevOps and QA teams on a mainframe-to-microservices transformation for Sabre's travel Global Distribution System.
- Established cloud-native delivery foundations on GCP, with CI/CD, infrastructure automation, observability and security scanning.
Technical Lead
2015 — 2021- Led cloud, SRE and DevOps strategy for Telia Group and other European telecom operators across Nordic markets, owning reliability, automation, security and cost planning.
- Ran SLO/SLA-driven operations using MTTA, MTTR and RPO/RTO practices across multi-country services, and owned client stakeholder management.
- Deployed releases on AWS and GCP (EKS, GKE, ECS, VMs) with Jenkins CI/CD, Ansible and shell automation; set up observability with ELK, Nagios, Prometheus and Grafana.
Tecnotree · Huawei Enterprise · QuadGen Wireless · Ericsson
2010 — 2015- Built foundational depth across OS and database administration, billing and mediation systems, telecom platforms and infrastructure operations.
Outcomes, not activity.
A few of the initiatives that moved the needle on cost, reliability and engineering productivity at enterprise scale.
AI Engineering at Enterprise Scale
Scaled AI developer tools to 30,000+ users, and a GenAI developer assistant adopted by 15,000+ engineers — spanning code review, test generation, documentation and knowledge retrieval.
Enterprise AI Knowledge Platform
Architected a production platform combining EKS, OpenSearch, semantic retrieval, reranking and LLMs to improve engineering knowledge discovery across 70+ platforms.
Cloud Cost Transformation
Delivered $1.8M+ monthly savings through cloud modernization, compute migration, data-transfer optimization and automated cost controls.
Engineering Automation
Automated infrastructure provisioning, access management and patch management — cutting operational processes from days or months to minutes or weeks.
Platform Modernization
Modernized identity, data and infrastructure platforms through SAML, Kubernetes, Presto and cloud-native architectures.
Awards & credentials.
Awards
- TechX Award 2026Capital One
- Icons of Excellence 2025Capital One
- Achievers Award 2024Capital One
- GenAI Hackathon 2024 — 2nd PrizeCapital One
- Hall of Fame 2023Paytm
- MVP Award 2019Mobileum
Education
- B.E. Electronics & Communication EngineeringB.R.C.M College of Engineering and Technology · Maharshi Dayanand University · 2005 — 2009
- Certified Kubernetes Administrator (CKA)Linux Foundation
Open to the next leadership challenge.
Reach out about Director / Head roles in Generative AI, AI platforms, platform engineering or developer productivity — or just to talk shop.