// Applications Engineer @ Oracle

Ishan Tewari.

View My Work ↗ LinkedIn
// 01 — About

A little about me.

Background

Engineer. Researcher. Guitarist.

  • Applications Engineer at Oracle — 3.5+ years in enterprise software
  • Migrated SOA → PL/SQL architecture, achieving 80% performance gain
  • B.Tech CS, Nirma University — 8.53 GPA
  • Deep understanding of system design at scale, microservices & HLD
  • Kaggle competitor, published researcher & book chapter author in AI/ML
  • Off-screen: guitarist, F1 fan & stand-up comedy enthusiast
Current employer 3.5+ Years at Oracle
Oracle
Academic 8.53 GPA / 10.0
Portfolio 6+ Projects Built
When I'm not coding
🎸 Guitar 🏎️ Formula 1 🎤 Stand-up Comedy 🎯 Carrom
// 02 — Experience

Where I've worked.

Oracle Applications Engineer Jun 2022 — Present Full-time
  • Led SOA → PL/SQL architecture migration, achieving 80% performance improvement across core business processes.
  • Resolved 150+ bugs across cross-functional teams, maintaining 100% regression-avoidance on production releases.
  • Achieved 100% regression avoidance through disciplined test coverage and peer review processes.
  • Mentored new hires during onboarding, accelerating ramp-up time on complex legacy systems.
Java ADF VBCS PL/SQL Oracle DB Jenkins REST APIs
Shrine Software Salesforce Developer Intern Jan 2022 — Jun 2022 Internship
  • Coordinated development with Mitsubishi Philippines client team, translating business requirements into Salesforce solutions.
  • Built 30+ Field Service Lightning (FSL) flows, streamlining field operations and reducing manual effort significantly.
  • Implemented UI components using LWC, Aura Components, and Visualforce pages within the Salesforce ecosystem.
LWC Aura Components Visualforce Salesforce FSL
// 03 — Projects

Things I've built.

Web App

Crisis Compass

Real-time pandemic dashboard with live global COVID stats, hospital resource booking flows, and multi-role authentication — built on Django and a public REST data layer.

Python Django REST
NLP

BharatBuzz

LDA-based topic modelling on Hindi/Hinglish tweet corpora, surfacing latent discourse themes with custom multilingual preprocessing and stop-word removal.

Python LDA NLP
Computer Vision

Flora Forensics

EfficientNetB7 fine-tuned on leaf imagery with aggressive data augmentation, scoring 81.7% accuracy — powering a Kaggle Top 151 finish worldwide.

Python EfficientNetB7 Deep Learning
NLP

MoodMapper

SVM classifier pipeline for large-scale tweet sentiment detection, achieving 73.4% accuracy — published as research at ACECAT 2021.

Python SVM NLP
Deep Learning

AlphaSignal

LSTM-based temporal model for equity price forecasting across multiple stocks, reaching 98.5% accuracy on held-out test sequences.

Python LSTM Keras
// 04 — Skills

My toolkit.

Languages

Java Python Swift C JavaScript PL/SQL

Frameworks

Java ADF VBCS Django Flask Keras Sklearn OpenCV Pandas

Databases

Oracle DB MySQL MongoDB PL/SQL

Tools

Git Jenkins Postman Salesforce LWC Aura Visualforce

Architecture

System Design Microservices High-Level Design SOA REST APIs Distributed Systems
// 05 — Achievements

Recognition.

Published Book Chapter

Co-authored a chapter on Skin-Colour Based Hand Segmentation Techniques published by IGI Global. Covers colour models, skin segmentation methods, and real-world applications in hand gesture recognition systems. Read chapter ↗

Kaggle Top 151

Plant Pathology 2021 AI Competition — ranked 151st worldwide out of thousands of competitors globally. Built EfficientNetB7-based pipeline for plant disease classification.

Published Research

ACECAT 2021 — "Sentiment Analysis using AI" accepted and published at an international conference, contributing to the growing body of research in applied NLP and machine learning.

Google Kickstart

Ranked 5,471st globally in Google Kickstart Round A — a competitive programming challenge that draws hundreds of thousands of participants from across the world each year.