01 — About

Building AI That
Ships at Scale

AI Engineer with 6+ years building production-grade LLM applications, agentic RAG systems, and autonomous agent workflows across AWS, Azure, and GCP.

Currently at Wells Fargo shipping LLM integrations processing 50K+ daily records, RAG pipelines, and LLMOps infrastructure that cut token costs by 35%.

Tech Stack

LLMs & Agents

Claude APIOpenAIAWS BedrockLangChainLangGraphCrewAIMCP

RAG & Vector

PineconepgvectorFAISSChromaEmbeddingsRe-ranking

Cloud & Infra

AWS (EKS, Lambda, S3)AzureGCPDockerKubernetesHelm

DevOps & IaC

TerraformGitHub ActionsArgoCDJenkinsAnsibleCDK

Backend & APIs

PythonFastAPINode.jsPostgreSQLRedisMongoDB

Observability

OpenTelemetryPrometheusGrafanaSplunk APMDatadog

02 — Experience

Where I've
Made Impact

Wells Fargo

AI Engineer

Feb 2024 — Present

Charlotte, NC

  • LLM integrations (Claude + OpenAI) classifying 50K+ daily records with 85%+ accuracy, cutting manual review by 60%
  • RAG pipelines over internal knowledge sources — 40% reduction in hallucination rates
  • Secure AI microservices (FastAPI, OAuth2, EKS) for 3+ autonomous agent workflows, saving 40 hrs/week
  • LLM observability with OpenTelemetry + Splunk APM across 12+ services — 50% lower MTTR, identified $80K+ avoidable token spend
  • 35% LLM API cost reduction via prompt compression, caching, model selection, and fallback routing
  • AIOps self-healing pipelines: incident response from 30 min → 8 min, prevented 15+ outages/year

Fluidra

AI Infrastructure Engineer

Jan 2023 — Dec 2023

Carlsbad, CA

  • AWS infrastructure for AI/ML workloads (EC2, ASG, S3, Lambda, EKS, CloudFront)
  • IaC with Terraform, Ansible, Boto3, CDK, and Serverless Framework
  • Intelligent monitoring and alerting — ~60% reduction in false alerts
  • CI/CD pipelines with Jenkins, ArgoCD, GitHub Actions across all environments
  • AWS cost optimization via right-sizing, unused resource cleanup, and IAM governance

Ericsson

DevOps Engineer

Jan 2019 — May 2022

India

  • CI/CD with Jenkins, Spinnaker, Kubernetes, Ansible, Puppet on GCP
  • Reusable Terraform modules for GCP and Azure infrastructure
  • Observability automation with Python, Ansible, New Relic, Sumo Logic, and PagerDuty

Auburn University at Montgomery

Graduate Teaching Assistant

Aug 2022 — Dec 2022

Montgomery, AL

  • Lab sessions, office hours, and grading for CS coursework
  • Mentored students on algorithms and software development fundamentals

03 — Projects

Things I've
Built

01

WhatsApp Personal AI Agent

Personal AI Assistant

Autonomous personal AI via WhatsApp using Twilio, FastAPI, LangChain, LangGraph, Claude, and OpenAI. Multi-model orchestration with MCP connectors, CrewAI multi-agent planning, and RAG over personal knowledge with vector embeddings.

Claude APIOpenAILangChainCrewAIFastAPITwilioRAGRailway
02

Cross-Border Investment & Tax Intelligence · Solo Build

DesiVest

Full-stack fintech SaaS for 4.5M+ US-based NRIs — the only tool unifying US and Indian brokerage portfolios into a single dual-currency dashboard. Deterministic Sell Simulator answers "if I sell today, what do I keep after both countries' taxes?" — computing US STCG/LTCG/NIIT, Indian §111A/§112A, and DTAA treaty relief with hold-longer suggestions, backed by 70+ regression tests at 100% branch coverage. Direct broker integrations via MCP + OAuth 2.1 (Robinhood, Zerodha Kite) with Fernet-encrypted token storage and a code-enforced read-only whitelist.

PythonFastAPINext.jsPostgreSQLMCPOAuth 2.1DTAAFIFO

04 — Contact

Let's Build
Together

Open to AI engineering roles, consulting, and interesting projects. I respond within 24 hours.

koukuntlarakesh2@gmail.com
(704) 506-1301GitHub

Rakesh Reddy Koukuntla · AI Engineer · 2026