Headline

Software Engineer @FinBox | Go | Java | Microservices | Distributed Systems | AWS |Data Structures & Algorithms | System Design

About

I'm a backend engineer at Finbox with 2+ years of experience, working on the infrastructure that powers lending integrations Over the last two years, I've taken end-to-end ownership of DocGen Engine — a config-driven document generation platform built in Go that automates RBI-mandated loan agreements, KFS documents, and KYC verification reports. What started as an internal tool now handles 4,000–8,000 requests per day, peaking at 14,700, with new lenders onboarding through a no code configs Beyond DocGen, I've built the webhook-mapping configuration dashboard that replaced a manual engineering task with a self-service UI, a multi-scheme encryption layer (RSA, AES-GCM, JWE) for secure client communication, and contributed to the GraphQL integration layer used across Finbox's API platform. I care about systems that are configurable, observable, and cost-efficient — and I try to build things that other engineers can use without needing to come back to me every time. Open to conversations about backend engineering, fintech infrastructure, and Go.

Location

Greater Bengaluru Area

AI stack

claude code

claude

codex

Work experience

Software Engineer Backend at FinBox, Sep 2024 - Present · 2 yrs 2 mos, • Designed and built the Document Generation Engine end-to-end, eliminating engineering dependency for lender onboarding and enabling self-service document publishing. • Architected an AST-based template introspection engine integrated with the Anthropic Claude API to automatically generate document field mappings. • Built a configurable webhook-mapping platform with spreadsheet-style configuration, bitmask-based rule matching, and deterministic rule resolution. • Developed a GraphQL integration layer and reusable workflow components that reduced backend complexity and improved request performance. • Integrated the Document Generation Engine with Finbox's API platform to reuse stored lender responses, eliminating redundant API calls during document generation. • Designed and implemented a Kafka → ClickHouse callback logging pipeline, providing a scalable, queryable audit trail for production webhook traffic. • Implemented secure lender integrations using RSA, AES, JWE, JWT, and mTLS with AWS SSM-based certificate and secret management. • Built and optimized scalable backend systems using Go, AWS, MongoDB, PostgreSQL, Redis, Docker, Kubernetes, and CI/CD pipelines. • Led production debugging and rollout of major lender integrations, trained 30+ engineers and managers, and became the top contributor to the team's repository with 1,202 commits. • Scaled production services handling 4,000–8,000 requests per day, with peak traffic reaching 14,700 requests/day.

Software Engineer Intern at FinBox, Apr 2024 - Aug 2024 · 5 mos

Education

Bachelor of Technology, Computer Engineering, Madanapalle Institute of Technology & Science

class 12, Mathematics, Narayana Junior College - India

Public links

github: https://github.com/Poojithkumar24

leetcode: https://leetcode.com/poojith_kumar