SOFTWARE ENGINEER ➔ AI ENGINEER

Abhijeet Hiwale

Buildingreliablesystems for AIand payments

Ibuildbackendmicroservices,paymentvalidations,andmemoryenginesthatrunstablyunderproductionloads.Nofillers,noAIslop.

Discuss a Contract Get Resume

2.8+

Years at Paycraft

95%

Transaction failure reduction

60%

P99 latency reduction

Feature velocity via AI agents

Python•

OpenAI•

LangChain•

PyTorch•

Hugging Face•

FastAPI•

Java•

Spring Boot•

React•

Next.js•

PostgreSQL•

MongoDB•

Redis•

AWS•

Azure•

Docker•

JavaScript•

Git•

Kubernetes•

Hibernate•

HTML5•

CSS3•

Python•

OpenAI•

LangChain•

PyTorch•

Hugging Face•

FastAPI•

Java•

Spring Boot•

React•

Next.js•

PostgreSQL•

MongoDB•

Redis•

AWS•

Azure•

Docker•

JavaScript•

Git•

Kubernetes•

Hibernate•

HTML5•

CSS3•

[Overview]

Core Competencies

I'm an AI-native Software Engineer based in Pune, specializing in Java Spring Boot microservices and agentic AI architectures. I bridge the gap between robust backend fundamentals and cutting-edge AI-augmented development.

Currently at Paycraft Solutions, I orchestrate scalable fintech solutions, having modernized enterprise expense systems and engineered high-throughput payment workflows. I love building systems that handle high concurrency without breaking a sweat.

My focus lies in building intelligent, autonomous systems and reliable memory engines for LLMs. I'm actively exploring new frontiers in AI agent workflows, vector databases, and multi-agent consensus pipelines.

I am currently open to new opportunities, including Full-Time roles and Contract positions. If you're building ambitious AI or backend products, let's connect!

Based in Pune, Maharashtra

Credentials & Training

Oct 2022 – Nov 2023

Full Stack Web Development Program

Masai School

Bengaluru, Karnataka

June 2018 – July 2021

Bachelor of Computer Applications

Christ College

Pune, Maharashtra

INFRA_MONITOR

ONLINE

Active systems

NSDL Bank EMS

Migrated to Spring Boot microservices with 60% response time improvement.

Ongoing

Gujarat Metro NCMC Integration

Airtel Payments Bank API with AES encryption for AFC systems.

Ongoing

RBL Bank Expense Management

Migrated to Spring Boot microservices with 60% response time improvement.

Ongoing

Wallet Top-Up Validation System

RBI-mandated limit enforcement reducing failures by 95%.

Ongoing

Legacy JSP → React Migration

30% page load improvement via component-based architecture.

8 months ago

Current Role:AI Native Software Engineer

Current Focus:Payments & AI Microservices

Availability:Open for Full Time & Contract roles

[Registry]

Selected Projects

Curated platforms and databases engineered to demonstrate core architectural patterns.

PROJECT 01

Sentinel AI-PR Review Agent

Autonomous code review via parallel multi-agent pipeline

An autonomous code review agent that analyzes GitHub PRs to proactively detect security vulnerabilities, code quality issues, missing tests, and documentation gaps.

→Engineered a parallel multi-agent pipeline using LangGraph, FastAPI, and Redis with fault-tolerant checkpointing.
→Implemented a deduplication aggregator with a Human-in-the-Loop (HITL) Next.js dashboard.
→Deployed a 5-service architecture via Docker Compose with dynamic LLM routing and strict budget guards.

PythonLangGraphFastAPINext.jsRedispgvectorDocker

Code

PROJECT 02

MemoryOS

Persistent memory engine for AI assistants

An API system that gives LLM agents reliable long-term memory. It allows them to store user preferences, update details when they change, and delete elements when requested.

→Uses PostgreSQL and pgvector for low-latency memory matching.
→Implements hybrid retrieval (BM25 + dense vector embeddings) for context accuracy.
→Maintains a p95 retrieval latency of 12ms under testing conditions.

PythonPostgreSQLpgvectorDockerRender

Code Live Demo

memory-os.api

memory_engine.py

"text-neutral-600 italic"># Deterministic memory retrieval
"text-amber-500 font-semibold">def "text-yellow-500">get_relevant_memories(user_id, query_embedding):
    "text-amber-500 font-semibold">return db."text-yellow-500">execute(
        "SELECT content FROM memories "
        "WHERE user_id = :uid "
        "ORDER BY embedding <=> :emb "
        "LIMIT 5",
        uid=user_id, emb=query_embedding
    )

PROJECT 03

Build.AI

AI-powered browser-based development platform

A browser-based development environment using the WebContainer API that enables full-stack application creation with real-time code execution, live previews, and one-click deployment — zero local setup.

→Integrated AI code generation with custom prompt engineering to automate component and API scaffolding.
→Real-time code execution and full-stack preview inside the browser via WebContainer API.

ReactJSNodeJSWebContainer APIAI Workflows

Code

build-ai.vercel.app

scaffold.js

"text-neutral-600 italic">// AI-powered scaffolding system
"text-amber-500 font-semibold">const "text-yellow-500">generateComponent = "text-amber-500 font-semibold">async (prompt) => {
  "text-amber-500 font-semibold">const scaffold = "text-amber-500 font-semibold">await ai.generate({
    template: prompt,
    context: projectContext
  });
  "text-amber-500 font-semibold">return webcontainer."text-yellow-500">mount(scaffold);
}

PROJECT 04

Redline Backend

Scalable microservices for a social discussion platform

A distributed microservices backend for a social discussion platform with 5+ services — User Management, Post Management, and Community Management — orchestrated via Spring Cloud.

→5+ microservices with Eureka, Config Server, and API Gateway for service discovery.
→95%+ API reliability with response times under 200ms with comprehensive error handling.

JavaSpring BootMongoDBSpring CloudEurekaAPI Gateway

Code

redline-backend.railway.app

PostService.java

// Eureka service registration
@EnableEurekaClient
@SpringBootApplication
public class PostService {
  // P99 < 200ms, 99.5% uptime
  public static void main(String[] args) {
    SpringApplication.run(PostService.class, args);
  }
}

[Metrics]

Production Feats

Low-latency architectures, transaction validations, and microservices engineered for critical financial infrastructure.

NCMC METRO PIPELINE

Gujarat Metro NFC Infrastructure

Built a microservice for secure Airtel Payments Bank API integration enabling card issuance, blocking, and tap-to-pay transaction processing for metro commuters. Encrypted all API payloads with secure AES-GCM-256.

ISSUED BY AIRTEL BANK

GUJARAT METRO NCMC CARD

XXXX XXXX XXXX 0789

AES_PAYLOAD_ENCRYPTED

VALIDATION ENGINE

Top-Up Validation

Developed a wallet top-up validation system implementing limit checks that reduced transaction failures by 95% and ensured NPCI compliance.

FAILURE REDUCTION-95%

RBI & NPCI COMPLIANT

LATENCY OPTIMIZATION

RBL Bank Migration

Led end-to-end migration of RBL Bank's Expense Management System to Spring Boot microservices, cutting response times by 60% with Redis caching and query optimizations.

P99 RESPONSE LATENCY-60% (800ms)

→

800ms

SYSTEM MONITOR

High-Volume Payment Flows

Engineered RESTful APIs for high-volume banking systems to support thousands of concurrent connections with zero downtime. Developed components in collaboration with cross-functional banking teams.

5,000+

Concurrent Users

30%

Cost Reduction

100%

Uptime SLA

[Timeline]

Work Experience

My professional milestones building payment microservices, transaction validations, and AI integrations.

Jan 2024 – Present

Software Engineer - I

Paycraft Solutions

Mumbai, Maharashtra

Fintech / Digital Payments

Owned and led the end-to-end migration of RBL Bank's Enterprise Expense Management System to a Java Spring Boot microservices stack, implementing stateless REST APIs, Redis-based caching, and query optimization to reduce response times by 60% (from 2s to 800ms). Built scalable services handling 5,000+ concurrent users with zero downtime during production rollout, leading to a 30% reduction in operational costs.
Led the migration of a legacy banking portal from JSP to React, introducing component-based architecture, optimizing API calls, and implementing state management techniques to enhance performance. Achieved a 30% improvement in page load times through systematic code refactoring and performance optimization.
Built RESTful APIs for payment processing workflows, handling high-volume transactions with optimized database queries and caching strategies. Adopted AI-agentic coding workflows to accelerate delivery of backend and frontend features, implementing user-facing React components and server-side APIs while meeting sprint deliverables and code quality standards.
Took end-to-end ownership of building scalable architecture for a regulated multi-tenant fintech platform. Engineered asynchronous batch processing workflows with data isolation for corporate clients. Integrated bank data providers and improved production reliability by resolving API latency bottlenecks.

JavaSpring BootMicroservicesRedisReactPostgreSQLREST APIsAWS

Oct 2023 – Jan 2024

Technical Assessor & Code Reviewer

Masai School

Bengaluru, Karnataka

EdTech / Tech Education

Evaluated and mentored 50+ engineers, conducting rigorous technical assessments and code reviews to align skills with production standards.
Analyzed software architecture assignments and provided high-impact, actionable system design feedback to developers.

Code ReviewMentorshipSoftware ArchitectureJavaJavaScript

Querying telemetry nodes...

[Connect]

Get in Touch

Start a Conversation

If you need help building payment integrations, setting up vector retrieval engines, or developing Spring Boot services, feel free to reach out.

Email: abhijeethiwalemail@gmail.com

Location: Pune, Maharashtra