HARSHA YELLELA
Software Engineer
Building scalable, cloud-native applications and AI-driven systems.
Currently at PNC Bank · M.S. Computer Science · Lawrence Technological University.
94
Lambda Functions
70%
Process Time Reduced
30
Projects Built
3 yrs
Professional Exp.
Enterprise tools used by Ferrari, Boeing & Triumph at Infor
Multi-agent AI pipelines deployed on AWS Fargate / EKS at LTU
Serverless SaaS (Resumade.in) with sub-2s generation & 95%+ uptime
Technical Skills
Tech Stack
Selected Work
Projects
Manual resume tailoring is slow and misses job-specific keywords.
◆ Outcome
80% less manual effort · sub-2s PDF generation · 95%+ uptime
Role: Sole architect & engineer
Lambda Microservices Platform
Serverless · Backend · Integrations
Enterprise SaaS needed 94+ isolated business-logic units with payments, comms, and e-signing.
◆ Outcome
94 serverless functions · real-time WebRTC calls · end-to-end payment flow
Role: Backend engineer
Telegram Toxicity Moderator
ML · Kubernetes · High Availability
Real-time moderation of high-volume Telegram channels at scale without over-provisioning.
◆ Outcome
500 RPS avg · 1 500 RPS peak · 70% infra cost reduction via spot instances
Role: ML + DevOps engineer
CRE Research Agent
RAG · Agents · NLP
Literature review across Semantic Scholar + ArXiv is manual and time-consuming.
◆ Outcome
Multi-source retrieval · embeddings + reranking · structured summaries in seconds
Role: AI engineer
Career
Experience
Software Engineer
PNC Bank
Jan 2026 – Present
USA
- Engineered scalable microservices using Java (Spring Boot) and Python to modernize fraud detection workflows, reducing transaction anomaly detection latency by 35%
- Architected an AI-driven RAG pipeline using LangChain and AWS Bedrock to enable contextual financial insights from structured and unstructured data
- Implemented event-driven data pipelines using Apache Kafka and Redis caching, reducing API response times below 200ms
- Containerized and deployed services on AWS EKS (Kubernetes + Fargate) to support peak loads of 50K+ daily transactions
- Reduced production downtime by improving observability with ELK stack, Prometheus, and Grafana, lowering MTTR by 40%
- Developed secure REST and GraphQL APIs with Node.js and Spring Security (OAuth2/JWT) for fine-grained access control
- Optimized CI/CD pipelines using Jenkins and GitHub Actions, accelerating release cycles by 30%
Software Engineer Research Assistant – Agentic AI
Lawrence Technological University
Jan 2025 – Dec 2025
Southfield, MI
- Designed multi-agent orchestration systems using CrewAI and LangChain to automate research workflows, reducing manual effort by 70%
- Built and deployed 3 persistent AI agent services on AWS EKS with Fargate, supporting semantic search across 10K+ documents with sub-second latency
- Developed FastAPI and GraphQL-based backend services integrated with AWS Bedrock for LLM-powered capabilities
- Implemented OpenSearch Serverless and Redis caching layers to optimize document retrieval and session management
- Automated infrastructure provisioning using Terraform and GitHub Actions CI/CD, reducing provisioning time from 2 hours to 15 minutes
- Conducted benchmarking between no-code (n8n) and coded (CrewAI) agent frameworks to guide architectural decisions
- Built scalable backend pipelines in Python with asynchronous processing to handle concurrent agent tasks
SDE-1 (LN Technical Consultant)
Infor India Pvt. Ltd.
Apr 2022 – Dec 2023
Hyderabad, India
- Developed enterprise-grade RESTful APIs using Java and Spring Boot to integrate ERP systems with external clients (Ferrari, Boeing, Triumph), processing 500+ daily transactions
- Engineered serverless microservices using AWS Lambda, API Gateway, and S3 triggers to automate data ingestion and event processing
- Diagnosed and resolved 15+ critical data pipeline issues across MySQL and distributed systems, reducing failure frequency from weekly to monthly
- Built backend services in C++ for performance-critical modules to optimize data processing and reduce latency in ERP integrations
- Containerized applications using Docker and implemented CI/CD pipelines, reducing release turnaround time by 25%
- Designed frontend components using React.js and JavaScript to improve user interaction with ERP dashboards
- Collaborated in Agile teams to deliver features across microservices and distributed architectures
Education
M.S. Computer Science
Lawrence Technological University
GPA: 3.77 / 4.0 · Dec 2025
B.Tech Computer Science & Engineering
Geethanjali College of Engineering
GPA: 7.5 / 10 · Aug 2022
Verified Credentials
Certifications
Get in Touch
Let's Connect
Software Engineer @ PNC Bank.
Building scalable cloud-native applications and AI-driven systems.