Software Engineer · TCS · India

Sayani Ghatak

Software Engineer building scalable web products and AI applications.

Software Engineer at Tata Consultancy Services with experience building production-grade frontend and full-stack applications. I also build independent AI-powered products using RAG, LangChain and modern web technologies.

Download resume
Sayani Ghatak, Software Engineer at Tata Consultancy Services

2+

Years at TCS

Banking products, BFSI

1st

Toycathon 2021

National winner, 1,000+ teams

9.25

B.Tech CGPA

JIS University

3X

TCS On-The-Spot Awards

2025–26

Selected work

Products I've designed, built and shipped.

Three products taken from idea to a deployed application — a PDF processing platform, a developer utility suite and an AI learning SaaS.

01Developer / Productivity Tool

PDF ToolBox

A single workspace for everyday PDF work — upload a document once, then organize, convert, secure or extract from it without switching between disconnected tools.

Common PDF tasks are scattered across a dozen single-purpose sites, most of which upload your file somewhere and keep it.

What I built

  • 31 tools across 7 categories in one workspace
  • Page organize, split, merge and crop on a single canvas with undo/redo
  • Watermark, page numbers, signature, redaction and form filling
  • AES-256 password protection and metadata stripping
  • Conversion both directions — Word, Excel, PowerPoint, HTML, CSV, JPG
  • OCR, text extraction and document translation
  • Files are processed and swept on a timer, never persisted to the database
TypeScriptReact 18ViteNode.jsExpressMongoDBTailwind CSSTanStack Querypdf-libmupdf
Screen recording of PDF ToolBox in useLive demo
02Developer Productivity

DevTool Kit

Developer utilities for everyday engineering tasks — 33 tools across JSON, encoding, debugging, API testing, diagrams and more, in one searchable workspace.

The usual answer is a bookmark folder of ad-heavy single-purpose sites, and you paste production data into every one of them.

33tools shipped
100%runs client-side

What I built

  • JSON formatter, validator and minifier
  • Base64 and URL encoders / decoders
  • JWT debugger, regex tester, diff checker, hash and timestamp tools
  • API client, mock API generator and OpenAPI doc export
  • SQL formatter, plus JSON ↔ CSV / YAML and XML → JSON conversion
  • Interactive DB schema and system design diagram builders
  • Every tool runs locally — pasted data never leaves the browser
Next.jsReactTypeScriptTailwind CSSRadix UILucide
Screen recording of DevTool Kit in useLive demo
03AI Learning Platform (SaaS)

SkillForge

A full-stack AI-powered SaaS learning platform featuring AI tutoring, mock interviews and smart quizzes.

Course platforms hand you video and hope it sticks — there is rarely anything that answers a question at the moment you get stuck.

60%less video buffering
45%better Lighthouse page speed

What I built

  • AI tutor that answers questions against the course content
  • AI mock interviews with real-time feedback
  • AI-generated study notes and adaptive quizzes
  • Structured learning paths with progress tracking
  • Video delivery through Mux with Vercel Edge caching
  • Sanity CMS studio for course and lesson management
Next.js 15React 19TypeScriptVercel AI SDKSanity CMSZustandClerkMuxTailwind CSS
Screen recording of SkillForge in useLive demo

Experience

Where I work.

TCS

Tata Consultancy Services

Software Engineer (Product Engineer)

BFSI Unit — Banking Products Team

Sep 2024 — Present

Kolkata, India

16+

client stakeholders presented to

10+

REST APIs integrated

60%

faster page loads

85%+

unit test coverage

~40%

fewer production bugs

  • Built the Consent Management System (CMS), DPDP Admin portal and DSR portal from the ground up using Angular and TypeScript to support enterprise-wide DPDP Act compliance; the product is under active evaluation and has been presented to 16+ client stakeholders.
  • Developed and maintained UI modules across multiple banking products — ULIP, SBI Loan, Business Partner, Product Configurator and Consent Management (DPDPA) — using Angular and TypeScript.
  • Integrated 10+ REST APIs across key modules (accounts, transactions, fund transfers), ensuring smooth and reliable data flow in production.
  • Improved application performance by reducing average page load time from 3.5s to 1.4s, a 60% improvement.
  • Increased unit test coverage to 85%+ by writing Karma and Jasmine test specs, reducing production bugs by ~40%.
  • Performed regular code refactoring and bug fixes to improve code quality and maintainability across multiple product modules.

JIS University

B.Tech in Computer Science & Engineering

CGPA

9.25 / 10

Years

Jun 2020 — Jul 2024

Toolkit

Tech stack.

Languages

JavaPythonTypeScript

AI / ML

LangChainRAG PipelinesGemini APIOpenAI APIPineconeVercel AI SDK

Databases

MongoDBMySQLPostgreSQLPrisma ORM

Frontend

React 19Next.js 15AngularTypeScriptTailwind CSSShadcn/UIHTML5CSS3

Backend

Node.jsExpress.jsREST APIs

Testing & Tools

KarmaJasmineGit

Engineering foundations

Fundamentals I keep sharp outside of work.

Data Structures & AlgorithmsSystem DesignObject-Oriented DesignDatabase DesignAPI Design

Currently exploring

DockerKubernetes

Also built

More things I've built.

Independent full-stack and AI projects. GitHub has the complete history.

ChatPDF

Upload a PDF and get context-aware answers in real time. Full RAG pipeline — chunking, Gemini embeddings, Pinecone storage and streaming responses — with auth, PDF viewer and chat history.

Next.jsTypeScriptLangChainGemini 2.5 FlashPinecone

SmartSite AI

Generates a complete website from a text prompt with live preview across mobile, tablet and desktop. Credit-based payments, secure login, inline editing and one-click publish.

ReactNode.jsExpressPostgreSQLPrisma ORMOpenAI API

Slack Clone

Real-time collaboration app built on the MERN stack and WebSockets, with threads, reactions, polls, video calls and file sharing.

MongoDBExpressReactNode.jsWebSockets

QuickGPT

Full-stack AI app for chatting with AI, generating images and managing your content in one responsive platform.

ReactExpressMongoDBImageKitOpenAI API

Research & recognition

Published work.

Machine learning and NLP research from my final year of study, cited 105+ times on Google Scholar.

IEEE Best Paper Award · 2023

An AI-based medical chatbot model for infectious disease prediction

IEEE Access

A deep feedforward multilayer perceptron and NLP architecture for predictive lifestyle medical chatbots, with a minimal testing loss of 0.1232.

View paper

105

Citations

2024

A study and comparison of deep learning based potato leaf disease detection and classification techniques using explainable AI

Multimedia Tools and Applications

43 citedView
2023

Voicebot For Mental Disease Prediction and Treatment Recommendation Using Machine Learning

Authorea Preprints

Integrates deep neural networks with NLP to build voicebot diagnostic frameworks for mental health screening.

4 citedView
2024

Public perception analysis on COVID-19 tweets using hybridised method

International Journal of Computational Biology

4 citedView
View Google Scholar profile

Awards & achievements

2023

IEEE Best Paper Award

International IEEE conference

Research paper on Machine Learning recognized as Best Paper at an international IEEE conference during my final year of study.

1st place

Toycathon 2021 — National Winner

All India

Ranked 1st among 1,000+ teams across India for designing a tech-based educational toy concept.

2025–26

3X On-The-Spot Award

Tata Consultancy Services

Awarded three times for consistently delivering high-quality work ahead of deadlines on a live banking project.

2024

B.Tech CSE — CGPA 9.25 / 10

JIS University

Graduated in Computer Science & Engineering with a 9.25 CGPA alongside published research.

About

A little about me.

I'm a Software Engineer at Tata Consultancy Services, on the BFSI banking products team. Most of my day is Angular and TypeScript — I built the Consent Management System, DPDP Admin portal and DSR portal from the ground up, and I work across UI modules for products like ULIP, SBI Loan and Product Configurator.

Outside of work I build full-stack products end to end: a PDF processing platform with 31 tools, a developer utility suite, an AI learning SaaS, and a few RAG applications using LangChain, Gemini and Pinecone. I like problems where the engineering is actually load-bearing.

I also have a research background in applied machine learning and NLP, including an IEEE Best Paper Award. I'm looking for engineering teams building scalable frontend systems, full-stack products or production-grade AI applications.

Sayani Ghatak

Currently

Software Engineer @ TCS

Building

Web products + AI applications

Exploring

System design · RAG · Scalable systems

Contact

Let's build something useful.

I'm open to software engineering opportunities where I can work on scalable frontend systems, full-stack products and production-grade AI applications.