Hey, Welcome to my digital space.

Sakshi
Kaveri

▶

Engineering scalable platforms, enterprise automation, and AI-powered products that turn complex business challenges into measurable outcomes.

View Projects Download Resume
2+
Years Experience
15+
Production Features
30%
Latency Reduced
99.9%
System Availability
scroll
// about

Who I Am

I'm a Software Engineer at Newgen Software, where I work on enterprise-grade backend systems for banking and BPM workflows. My day-to-day revolves around Java, Spring Boot, and REST APIs — keeping systems fast, secure, and highly available.

Beyond my primary stack, I'm deeply curious about AI engineering. I've built production-grade RAG pipelines, worked with vector databases, and prototyped LLM-based document extraction systems. I believe the next decade of backend work will be inseparable from AI.

I hold a BTech in Electronics & Telecommunications with a 9.11 CGPA — I pick things up fast and go deep once I'm in.

// education
BTech — Electronics & Telecommunications
Don Bosco Institute of Technology · May 2024
CGPA 9.11 / 10
// currently
● Newgen Software · Software Engineer
Feb 2024 – Present
Fintech · BPM · Enterprise Backend
// experience

Where I've Worked

Feb 2024 – Present
Newgen Software
Full-time

Software Engineer

  • End-to-end backend engineering for enterprise banking workflow features using Java, Spring Boot, JSP, and REST APIs. Implemented secure RBAC and authentication workflows.
  • API design & integration: Designed REST API integrations for banking workflow applications and maintained SOAP-based integrations for legacy systems.
  • Optimized backend logic, database queries, and connection pooling — reducing latency by 30%perf win while sustaining 99.9% availability.
  • Delivered 15+ production features from design through deployment with full ownership and post-release support.
  • CI/CD pipeline setup with GitHub Actions, automated testing, and zero-downtime deployments on Tomcat and JBoss.
  • AI prototyping: Explored LLM-based document extraction and automation approaches for BPM document workflows.
// skills

Technical Stack

// languages
Java Python SQL JavaScript
// backend & frameworks
Spring Boot Spring MVC Spring Security Hibernate / JPA FastAPI RESTful APIs Microservices
// databases
PostgreSQL MySQL Oracle PL/SQL pgvector
// ai / llm
RAG Pipelines Vector Embeddings Semantic Search GPT-4o-mini Groq LLM Sentence Transformers
// devops & tools
Git GitHub Actions Docker Maven Postman Tomcat JBoss
// concepts
OOP DSA Auth & RBAC CI/CD DB Optimization JWT
// projects

Things I've Built

🔍
AskMyDocs
AI-Powered Document Q&A Platform

A document Q&A assistant enabling natural language search and querying across large document collections using semantic retrieval and LLMs.

  • End-to-end RAG pipeline with embeddings, pgvector indexing & source-cited responses
  • Sub-100ms retrieval via IVFFLAT indexing + cosine similarity
  • Semantic search over arbitrary document collections
Python FastAPI PostgreSQL pgvector Sentence-Transformers RAG
View on GitHub →
⚡
Extractly
LLM-Based Document Extraction API

An API that converts unstructured invoices, receipts, and emails into structured JSON for downstream business workflows at scale.

  • Llama 3.3 70B via Groq API with Pydantic schema validation
  • 95%+ extraction accuracy across multiple document formats
  • Fully containerized with Docker for consistent deployment
Python FastAPI Groq API Llama 3.3 70B Pydantic Docker
View on GitHub →
🔐
Auth & Access Control API
Centralized Authentication Service

A centralized authentication and authorization service for securing backend applications, with JWT auth, RBAC, and full CI/CD automation.

  • JWT authentication + RBAC + encrypted credential storage
  • GitHub Actions CI/CD with cloud deployment
  • Automated Postman test suites for release reliability
Spring Boot Spring Security MySQL REST APIs GitHub Actions JWT
View on GitHub →
// contact

Let's Build Something Great

I'm currently open to Software Engineering, AI Engineering, and Backend Development opportunities.