Backend architecture
REST APIs, async Python services, multi-tenant data models, validation, migrations, and release-ready code.
Python backend + AI automation
Hi, I am Shivam Gupta.
Backend engineer with 2+ years building FastAPI services, async ingestion pipelines, AWS Lambda jobs, and OpenAI-powered reconciliation for custodian files, portfolio transactions, QSR feeds, and enterprise treasury workflows.
About
My work sits where financial files meet production backend systems. I design FastAPI endpoints, model tenant-aware PostgreSQL schemas, build daily Python automation on AWS, and turn custodian/QSR transaction files into reliable platform data.
REST APIs, async Python services, multi-tenant data models, validation, migrations, and release-ready code.
OpenAI API reconciliation, prompt-driven validation, and AI-assisted debugging, test generation, and documentation.
AWS Lambda, SQS, S3, CloudWatch, Inspector remediation, and scheduled jobs for daily operational runs.
Selected work
Built a Python reconciliation workflow that checks daily securities holdings from custodian banks against internal balance-sheet data and flags mismatches before manual review.
Architected CUSIP-level ingestion and query APIs across CPs, T-Bills, Bonds, Agencies, and Corporate Bonds for enterprise treasury clients.
Created Python Lambda functions and SQS integrations for recurring financial processing, scheduled validation, service monitoring, and cloud hardening.
Experience
Building FastAPI microservices, async ingestion workflows, OpenAI-powered reconciliation, AWS Lambda jobs, and tenant-aware PostgreSQL systems for enterprise treasury clients.
Authored REST APIs across Spring Boot and Quarkus microservices, implemented OAuth2 authentication, worked with Kafka and Redis, and shipped backend updates through bi-weekly AWS releases.
Built a Selenium automation framework with the PageFactory pattern, covering 30+ end-to-end test cases and reducing manual QA effort for an ESG application.
Completed structured enterprise training in C#, ASP.NET MVC, and web application development through hands-on projects.
Projects
Python / TensorFlow / Flask
Built a Bi-LSTM next-word prediction system trained on 20,000+ sequences, reaching 91% accuracy and serving real-time predictions through a Flask interface.
Java / Maven / PostgreSQL
Authored a schema comparison library that detects drift across Dev, QA, and Prod PostgreSQL environments and generates SQL reconciliation scripts.
Python / OpenCV / Dlib / Flask
Implemented a real-time computer vision pipeline using EAR/MAR thresholds and Dlib/OpenCV face detection, processing roughly 25 frames per second with alerting.
Skills
Signals
Contact
I am open to Python backend, AI engineer, and cloud automation roles where clean APIs, strong data validation, and pragmatic AI workflows matter.