- 100K+
Built a Go gift-card campaign rules engine that processes 100K+ transactions a day, checking conditions and eligibility rules within 150ms.
- −40%
Connected the rules engine to the coupon service over gRPC and Protocol Buffers, cutting inter-service overhead by 40% versus the old API integration.
- −30%
Sped up rule processing by compiling knowledge bases outside critical sections and batch-loading 1,000+ rules per scope, reducing API response time by 30%.
- 99.5%
Designed REST APIs for campaign setup, rule validation, transaction processing and redemption, serving 500K+ requests a day at 99.5% uptime.
- −65%
Added Redis caching for service tokens and campaign data with an 85% hit rate, cutting downstream service calls by 65%.
- 1K+
Built concurrent bulk upload with goroutines and buffered channels, handling 1,000+ records per upload with duplicate detection and field-level validation.
PuspitaSamasi.
Backend engineer who writes Go for payments: rules engines, gRPC services and high-concurrency APIs.
3+ years building Golang microservices for gift-card and payment workflows. I design systems that evaluate thousands of business rules in milliseconds, stay up under half a million requests a day, and use Generative AI tools to make the development cycle faster.
- Go
- gRPC
- Kafka
- Redis
- MySQL
- AWS
- Docker
- txn#48213ProductCodePASS12ms
- txn#48214PaymentCodeBLOCK9ms
- txn#48215CampaignWindowPASS14ms
- txn#48216MaxAmountPASS11ms
- 100K+gift-card transactions evaluated daily
- 500K+API requests a day at 99.5% uptime
- <150msto evaluate a transaction against its rules
- 85%Redis cache hit rate on campaign data
Three years on one payments platform, from intern to owner of the campaign engine.
- Wrote flexible, reusable code and libraries for upcoming product features.
- Shipped features and fixed bugs as a backend developer on the product team.
- Built APIs to improve project integration, working in PHP and Magento 2.
- Worked alongside designers, developers and system administrators.
Try a transaction against the rules.
A simplified version of the eligibility checks the campaign engine runs. Pick the product, payment gateway, amount and day, and watch each condition pass or block the transaction.
rule "Transaction" salience 10 { when Txn.ProductCode == "amazon" && Txn.PaymentCode == "payu" && Txn.Amount <= 10000 && Campaign.IsActive(Txn.Time) then Txn.ApplyCashback(10); Retract("Transaction"); }
Side work: fraud rules and recommendations.
Transaction Risk Management System
A fraud-prevention engine built on velocity and amount limits. Store and product limits enforce hourly and monthly thresholds, and dynamic rule evaluation blocks transactions that go over them. It plugs into an RFM transactions pipeline to flag risky behavior and anomalies in real time.
Movie Recommendation System
Content-based filtering that predicts what a viewer will like from their history, using movie metadata such as genre, ratings and keywords to produce personalised recommendations.
What I build with.
Computer applications, built on a maths foundation.
NIT Jamshedpur
Master of Computer ApplicationsSep 2020 – May 20238.1CGPAGovernment Autonomous College, Rourkela
B.Sc. Mathematics & Computer ScienceSep 2017 – Jul 20208.3CGPA- ★Recognised at Pine Labs for outstanding contributions to a high-impact project.
- ◆Member of the Competitive Programming Wing, NIT Jamshedpur.
Let's build something fast.
Open to backend and platform roles working with Go, distributed systems and payments. The quickest way to reach me is email.