Open to roles · Remote ok

Koushik

Gudipelly

Backend · Fullstack · ML engineer with 4+ years building fraud detection platforms, real-time streaming pipelines, and distributed microservices at Bank of America and Deloitte.

PythonFastAPIApache KafkaSparkNext.jsKubernetesAWS · Azure
See my workDownload resume
Fraud pipeline · live at BofA
txn/s 847lat 851ms
TXNKAFKASPARKXGBSTAPI
Kafka · 3 topics28 features<1s latency95.2% accuracy
M
transactions / day
+
years experience
%
deployment cycle cut
%
earlier fraud detection
The road so far

Journey

WorkAug 2024 — Present

Software Engineer, Backend / Python

Bank of America·NC, USA

Building a real-time fraud detection platform that scores 1–2M transactions a day and blocks fraud before settlement.

Sub-second scoring+12–15% accuracyKafka + Spark−30% deploy time
EducationAug 2023 — May 2025

M.S. in Computer Science

NC State University·Raleigh, NC, USA

Deepened distributed systems, machine learning, and data-engineering foundations while working in industry.

Distributed systemsMachine learningData engineering
WorkJan 2021 — Jul 2023

Full Stack Software Developer

Deloitte·Hyderabad, India

Built FastAPI microservices for an e-commerce order platform at 40–50K daily transactions and led a monolith-to-microservices migration.

Monolith → microservicesREST + GraphQL APIsKafka + WebSocketsRedis caching
Selected work

Projects

Professional · Bank of AmericaBackend · ML · Streaming

Real-time fraud detection pipeline

Architected a real-time fraud detection platform at Bank of America processing 1–2M daily transactions across card payments, Zelle, ACH, and online banking. Built event-driven Kafka + Spark pipelines, ML-backed XGBoost scoring, rule-based detection logic, and FastAPI microservices — blocking fraud within 1–2 seconds before settlement.

Transaction input→Kafka topic→Spark Streaming→Feature engineering→XGBoost scorer→FastAPI response
1–2M
daily transactions
<1s
scoring latency
15%
earlier detection
30%
faster deploys
PythonFastAPIApache KafkaSpark StreamingXGBoostScikit-learnMySQLDockerKubernetesJenkins
Read case study →Architecture diagram
Side projects
🚦
Backend · Infra

Distributed rate limiter

FastAPI gateway enforcing distributed rate limits with Redis Lua scripting, fail-open/fail-closed modes, Prometheus observability, and Docker Compose deployment.

PythonFastAPIRedisDockerPrometheus
GitHub
🔍
AI · Fullstack

AI resume analyzer

Upload a resume and job description — get instant match scoring, gap analysis, and rewrite suggestions via Gemini API.

ReactFastAPIGemini APIPython
Live demoGitHub
📊
Fullstack

Project management platform

Full-stack app with Gantt charts, dashboards, Cognito auth, and Prisma ORM deployed on AWS EC2 and RDS.

Next.jsTypeScriptNode.jsAWS
Live demoGitHub
Capabilities

Skills

Backend and APIs
FastAPI / PythonMicroservicesSpring BootGraphQL / RESTAsync programming
Data and streaming
Apache KafkaSpark / PySparkPostgreSQL / MySQLRedis / SnowflakeETL pipelines
Cloud and DevOps
Docker / KubernetesAzure AKSAWS (EC2/S3/RDS)Jenkins / GitLab CIPrometheus / Grafana
ML and analytics
XGBoost / Scikit-learnTensorFlow / CNNsFeature engineeringPandas / NumPyAnomaly detection
Also worked with
PythonJavaTypeScriptJavaScriptSQLFlaskWebSocketsPydanticOpenAPIJWT / OAuth2RBACSonarQubePyTestMockitoAWS CognitoAWS GlueAWS AmplifyNexusPrisma ORMPySpark

Let's build something.

Open to backend, fullstack, and ML-adjacent roles. Open to full time, remote and hybrid.

EmailLinkedInGitHub
Get in touch →