Backend Software Engineer & LLM Engineer

Sanjay Kandimalla

Charlotte, NC ********************@*****.*** 682-***-****

SUMMARY

• Software Engineer with 4+ years of experience building and scaling backend systems, APIs, and machine learning applications for Fortune 500 companies in aviation and financial services.

• Earned an M.S. in Applied Statistics and Data Science and applied it to production machine learning pipelines and LLM applications, including RAG systems, agentic workflows, and prompt engineering.

• Delivered measurable improvements in system performance, reliability, and cost across high-throughput backend services processing millions of daily transactions and requests.

• Deployed LLM-powered tools using LangChain, LangGraph, and the Claude and OpenAI APIs, applying structured prompt engineering and retrieval architectures to production use cases.

• Owned projects end-to-end, spanning data pipeline design, model development, API deployment, and production monitoring.

• Converted ambiguous business problems into scalable, tested, production-ready technical solutions. CORE COMPETENCIES

Python System Design Backend Engineering API Development Microservices Distributed Systems Machine Learning LLM Engineering Agentic Workflows RAG Vector Search Prompt Engineering Statistical Modeling Data Pipelines Cloud Infrastructure CI/CD DevOps

TECHNICAL SKILLS

Languages: Python, SQL, Java, JavaScript, TypeScript, Bash Backend/API: FastAPI, Flask, Django, REST, GraphQL, gRPC, Microservices, Event Driven Architecture, WebSockets AI/ML/LLM: PyTorch, TensorFlow, scikit-learn, Pandas, NumPy, LangChain, LangGraph, LlamaIndex, CrewAI, OpenAI API, Anthropic Claude API, Hugging Face Transformers, RAG Architecture, Agentic Workflows, Prompt Engineering, Fine-Tuning (LoRA/PEFT), Model Context Protocol (MCP), LLMOps, Model Evaluation

Data & Statistics: Statistical Modeling, Hypothesis Testing, A/B Testing, Regression, Time Series Forecasting, Feature Engineering, Experimentation Design, Bayesian Methods

Data Stores: PostgreSQL, MySQL, MongoDB, Redis, DynamoDB, Pinecone, FAISS, ChromaDB, Weaviate Cloud & Infra: AWS (EC2, S3, Lambda, SageMaker, RDS, Bedrock), Azure, GCP, Docker, Kubernetes, Terraform Engineering Practices: CI/CD (Jenkins, GitHub Actions), TDD, PyTest, Kafka, RabbitMQ, Design Patterns, Datadog, Grafana, System Design, Code Review

EXPERIENCE

Backend Engineer, Python American Airlines February 2025 – Present

• Architected end-to-end ownership of 12+ Python microservices (FastAPI/Flask) from design through deployment, serving 2M+ daily requests at 99.95% uptime for flight-operations and customer-facing platforms.

• Delivered high-throughput RESTful and GraphQL APIs by collaborating directly with product and data engineering teams, cutting cross-system data exchange latency by 35%.

• Rebuilt the core PostgreSQL/MySQL database layer through indexing and Redis caching, reducing query response time by 40% and infrastructure spend by 20%.

• Integrated predictive ML models for flight-delay and demand forecasting into production backend systems, partnering with the data science team to lift forecast accuracy by 18%.

• Containerized and orchestrated services using Docker and Kubernetes, compressing deployment time from hours to under 15 minutes.

• Launched an internal LLM-powered assistant using the Claude API, LangChain, and a ChromaDB-backed RAG pipeline, trimming support-ticket resolution time by 25%.

• Engineered agentic workflows with LangGraph to automate multi-step operational tasks, shrinking manual processing time by 30%.

• Applied structured prompt engineering, including few-shot examples and chain-of-thought reasoning, raising LLM output reliability by 20%+.

• Expanded automated test coverage to 90%+ across owned services, driving a 50% drop in production incidents.

• Established observability pipelines (Datadog/Grafana) in partnership with the platform reliability team, shortening mean time to resolution by 35%.

• Migrated legacy monolithic services to a microservices architecture, cutting long-term scaling costs by 25%.

• Standardized API documentation (OpenAPI/Swagger) across 10+ services, shortening partner-team onboarding time by 30%.

• Conducted load testing ahead of peak travel season, confirming system stability under 3x normal traffic.

• Evaluated LLM output quality using automated evaluation frameworks, improving model selection accuracy by 15%.

• Presented architecture proposals to senior engineering leadership, directly shaping platform-wide technical direction.

• Mentored 2 junior engineers through code reviews and system design sessions, boosting team sprint velocity by 15%. Python Developer American Express June 2021 – December 2023

• Developed backend services and ETL pipelines supporting fraud-detection and risk-analytics platforms processing 5M+ transactions daily.

• Streamlined core data pipelines using Pandas and SQL, trimming processing time from 6 hours to under 2 hours.

• Partnered with the data science team to productionize ML classification and anomaly-detection models as scalable REST APIs, improving fraud-detection accuracy by 22%.

• Shipped internal platforms and APIs (Flask/Django), collaborating across 5+ cross-functional teams and eliminating 10+ hours per week of manual reporting work.

• Implemented automated monitoring and alerting, curbing mean time to detection for production issues by 40%.

• Performed statistical modeling and hypothesis testing to validate model performance, supporting risk and fraud decisions made jointly with business stakeholders.

• Contributed to early NLP experimentation for transaction-note classification, laying the technical foundation for later automation efforts.

• Raised unit and integration test coverage to 85%+, reducing post-release defects by 30%.

• Optimized SQL query and indexing strategy across core transactional databases, lowering average query time by 30%.

• Automated data validation and reconciliation checks, decreasing manual QA effort by 8+ hours per week.

• Authored technical runbooks and documentation across 6+ services, shrinking new-engineer onboarding time by 20%.

• Coordinated with compliance and security teams to enforce data-handling standards, strengthening protection of sensitive financial data.

• Refined CI/CD pipelines in Jenkins, minimizing average build and deployment time by 35%.

• Guided 3 cross-functional stakeholder groups through sprint reviews and technical demos, curbing requirement-related rework by 15%.

• Owned production support rotations across 20+ Agile sprints, maintaining platform reliability during high-volume transaction periods.

• Drove adoption of code review standards across the team, enhancing code quality consistency among 4 fellow engineers. PROJECTS

Agentic RAG Assistant with Multi-Source Retrieval (Independent Project) Designed an agentic AI assistant combining LangGraph-orchestrated multi-step reasoning with a hybrid RAG pipeline (ChromaDB and BM25 keyword search) to answer complex queries across 10,000+ technical documents and structured data sources. Configured the agent to decide dynamically when to retrieve documents, query a SQL database, or call external APIs, improving answer relevance to 90%+ and cutting hallucination rate by 40% compared to a single-shot RAG baseline. Deployed as a FastAPI service using the Claude and OpenAI APIs, containerized on AWS, sustaining 500+ requests per second at sub-100ms latency. EDUCATION

M.S., Applied Statistics and Data Science University of Texas at Arlington CGPA: 3.9 December 2025 ACHIEVEMENTS

2nd Place, Technological Business Hackathon conducted by AIESEC November 2020

Contact this candidate