Skip to content
View shreyasshelar's full-sized avatar

Block or report shreyasshelar

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
shreyasshelar/README.md

Shreyas Shelar

Software Engineer  ·  Distributed Systems  ·  Fintech  ·  Cloud

Pune, India


LinkedIn   Email   AWS Certified



Backend engineer building production-grade distributed systems at scale. At Kanaka Software, I engineer a forex treasury management platform serving 16,000+ clients across swaps, spot, tom/cash, and forward contracts — owning real-time trade execution, post-trade settlement, and platform-wide infra services.


What I Build

  • High-throughput pipelines — Reduced a data ingestion job processing 81K API calls from ~269 days → 6 hours (>99.9% improvement) via nested parallelism and HTTP/1.1 upgrade
  • Distributed backends — Event-driven microservices, Saga orchestration, CQRS, transactional outbox, distributed caching — engineered for correctness under failure
  • Fintech infrastructure — Payment systems, ledger design, reconciliation engines, fraud detection pipelines, OAuth2/OIDC multi-IdP federation, AES encryption at rest
  • Greenfield services — Notification engine, email pipeline, distributed Quartz scheduler, document management system (1,000+ encrypted docs/month), custom DB-to-DB integration engine replacing AWS DMS

Flagship Project

AegisPay — Event-Driven Fintech Platform

Production-grade payment infrastructure built from the ground up.

Java 21 Spring Boot 3.3 Apache Kafka PostgreSQL ClickHouse Redis Kubernetes ArgoCD AWS EKS

  • 10 microservices over 18 Kafka topics (KRaft) with 5-step Saga orchestration, compensating transactions, and exactly-once delivery via transactional outbox
  • Stripe PaymentIntents + 3DS, multi-IdP OAuth2/OIDC (Keycloak / Entra / Okta), zero-trust JWT relay, dual-key Redis token-bucket rate limiting
  • Spring Batch reconciliation comparing ledger COMMITs against Stripe Balance Transactions API — catching MISSING_IN_STRIPE, MISSING_IN_LEDGER, AMOUNT_MISMATCH
  • 4 AI components — RAG fraud copilot (pgvector + Claude), agentic incident triage, multimodal KYC/OCR, LLM error resolution — all with AI audit log for RBI/DPDP compliance
  • GitOps on AWS EKS via ArgoCD, umbrella Helm chart across 3 environments, GitHub Actions parallel CI/CD, secrets via HashiCorp Vault + ESO, full distributed tracing (W3C → Micrometer → Prometheus → Grafana)

Tech Stack

Backend & Architecture

Java Spring Boot Spring Security Spring Batch Apache Kafka REST APIs

Databases & Streaming

PostgreSQL Redis MongoDB ClickHouse BigQuery pgvector

Cloud & DevOps

AWS GCP Kubernetes Docker Helm ArgoCD GitHub Actions

Observability & Security

Prometheus Grafana HashiCorp Vault OAuth2/OIDC

AI & LLM

Claude API RAG Agentic Workflows


By the Numbers

16,000+ clients on the forex treasury platform
>99.9% pipeline runtime reduction — ~269 days → 6 hours
81,000 API calls re-engineered with nested parallelism + HTTP/1.1
800ms → 200ms p95 API latency after N+1 fix + composite indexes + distributed cache
1,000+ AES-encrypted documents delivered per month
5 greenfield services built and owned end-to-end
10 microservices in AegisPay with exactly-once delivery guarantees
1 AWS certification — Data Engineer Associate, Oct 2025

Certification

AWS Certified Data Engineer — Associate  ·  Amazon Web Services  ·  Oct 2025


Pinned Loading

  1. AegisPay AegisPay Public

    Production-grade event-driven fintech platform — 10 microservices, Saga + CQRS, Kafka, AWS EKS, GitOps, AI fraud copilot