★ best viewed in Netscape Navigator 4.0+ ★ 800×600 resolution recommended ★ hand-coded with vim & too much chai ★ ★ best viewed in Netscape Navigator 4.0+ ★ 800×600 resolution recommended ★ hand-coded with vim & too much chai ★ ★ best viewed in Netscape Navigator 4.0+ ★ 800×600 resolution recommended ★ hand-coded with vim & too much chai ★ ★ best viewed in Netscape Navigator 4.0+ ★ 800×600 resolution recommended ★ hand-coded with vim & too much chai ★ ★ best viewed in Netscape Navigator 4.0+ ★ 800×600 resolution recommended ★ hand-coded with vim & too much chai ★ ★ best viewed in Netscape Navigator 4.0+ ★ 800×600 resolution recommended ★ hand-coded with vim & too much chai ★
vol. iv — issue 26all caps = enthusiasm

The Rahul Parida
Quarterly

a hand-stitched dossier of one engineer’s love affair with SaaS, AI agents, and the impossible bug

— greeting from the editor —

hi, i’m rahul.
i build SaaS things,
then ask why don’t they exist already?

I’m a final-year Computer Science student at Siksha ‘O’ Anusandhan University, Bhubaneswar — perpetually convinced that the next problem I stumble into deserves its own product. My favourite kind of evening is one where I’ve gone from “huh, that’s annoying” to a working prototype by 2 AM, and my favourite kind of morning is the one where someone actually pays for it.

On this site you’ll find a curated, slightly opinionated tour of the products I’ve built, the internships where I learned how the grown-ups do it, and the things I’m tinkering with right now. It is intentionally a little handmade. sorry, not sorry.

chapter one

About the editor

I’m a final-year CS undergrad at Siksha ‘O’ Anusandhan University, Bhubaneswar (CGPA 8.5), and I spend most of my waking hours somewhere between a terminal and a whiteboard. I love the moment a vague complaint turns into a working prototype — usually over a weekend, usually with too much coffee involved.

My current obsessions are agentic AI systems and the unglamorous backend plumbing that makes them reliable in production — think Kafka pipelines, Elasticsearch queries that don’t fall over at 3 AM, and circuit breakers that earn their keep. The projects further down this page are the receipts. — turn the page for the goods —

from the desk of

What I’m up to, lately

● LIVE · refreshes whenever i remember
  • readingDesigning Data-Intensive Applications (re-read)
  • shippingKnowBot v2 — adding agentic eval suite
  • studyingKubernetes the Hard Way, ch. 4
  • drinkingfilter coffee, third cup, slightly burnt

chapter two

The workbench — products i’ve shipped

Six small product-shaped things I’ve built, mostly by turning a real annoyance into a weekend project. Some grew up into internship work, some are still sitting on my hard drive judging me.

PROJ #01 · 2026SHIPPED

LogSearch

Distributed Log & Ticket Intelligence

Production-grade analytics platform for high-volume logs and support tickets. Built around an event-driven microservice backbone with a soft spot for circuit breakers.

  • Spring Boot microservices ingesting logs & tickets through Kafka topics
  • Elasticsearch with fuzzy search, autocomplete, and Redis caching at the edge
  • Resilience4j circuit breakers so one bad pod doesn't take the building down
  • JUnit5 + Mockito + Testcontainers wired into CI/CD for every PR
JavaSpring BootKafkaElasticsearchRedisPostgreSQLDocker
PROJ #02 · 2026SHIPPED

KnowBot

Multi-Agent Enterprise Knowledge Assistant

A team of small AI agents (planner, retriever, reranker, generator, critic) that answer questions about messy enterprise wikis with grounded citations.

  • Hybrid retrieval: vector search + BM25, with semantic caching for repeat questions
  • Confidence-based guardrails so the bot politely declines rather than hallucinates
  • Built an eval harness measuring faithfulness & relevance to debug answer quality
  • Front-end in React with citations panel & diff view of what changed
PythonFastAPILangGraphElasticsearchVector DBReact
PROJ #03 · 2025SHIPPED

Ticketer

AI-Driven Support Ticketing

Turns a flood of incoming tickets into a tidy queue: auto-categorised, prioritised, and routed to the right human.

  • Role-based access: user, moderator, admin — gated by JWT
  • Background jobs via Inngest for heavy LLM work that shouldn't block the UI
  • LangChain-driven summaries & suggested replies for moderators
  • Honest, useful AI features instead of AI-for-AI's-sake
Next.jsTypeScriptTailwindLangChainInngest
PROJ #04 · 2024SHIPPED

AnganSakhi

Field-Worker App for Anganwadi Centers

A mobile app that helps Anganwadi workers track children and mothers, with growth predictions built on UNICEF standards.

  • Custom algorithm using UNICEF reference data for height/weight forecasting
  • Scalable Express backend, deployed and in active use by real workers
  • Offline-friendly flows because field internet is… optimistic
React NativeExpress.jsTypeScriptPostgreSQLTailwind
PROJ #05 · 2024SHIPPED

TalkATive

Real-Time Chat, with Presence

A full-stack chat app: instant messages, online indicators, throwaway rooms. Basically the project that taught me websockets aren't magic.

  • Live presence tracking with online / offline indicators
  • Dynamic rooms — create, invite, leave, rejoin
  • Responsive UI that actually works on a phone in 2024
ReactExpressMongoDBSocket.IO
PROJ #06 · 2024SHIPPED

MealMate

Food Service, Web-First

A clean little food-ordering platform. MVC on the backend, GSAP for the flourish, Razorpay for the boring-but-important part.

  • Full CRUD for menus, profiles, and orders — modular MVC structure
  • Secure payments via Razorpay, including the ugly webhook parts
  • GSAP animations sprinkled like seasoning, not poured like ketchup
ReactExpress.jsTailwindGSAPRazorpay

chapter three

Where i’ve worked, briefly

  1. Refold.ai

    Sept 2025 – Nov 2025

    Web Developer Intern · remote

    • ·Shipped high-converting marketing & sales landing pages in Next.js + TypeScript + Tailwind.
    • ·Built & maintained internal dashboards and a small library of reusable UI components.
    • ·Spent more time than I expected in code reviews — and I’m grateful for it.
    01
  2. Samsung R&D Institute

    Dec 2024 – July 2025

    Research Intern · remote

    • ·Worked with a cross-functional team to collect & annotate 10,000+ images for YOLOv8 training.
    • ·Built the Python backend APIs and a small mobile app that consumed the trained model.
    • ·My first taste of what ‘end-to-end AI system’ actually means in production.
    02
  3. BLEM

    June 2024 – July 2024

    Web Developer Intern · remote

    • ·Built a custom, scalable CRM in Next.js + TypeScript + PostgreSQL + Prisma.
    • ·Wired up Facebook API for automated lead capture — and finally understood OAuth at a deeper level.
    • ·Designed & debugged the responsive UI across desktop & mobile.
    03

chapter four

The workshop — tools on the bench

languages i write in

JavaPythonTypeScriptJavaScriptC++CSQL

things i build with

Spring BootNode.jsExpress.jsFastAPIReactNext.jsReact NativeTailwind

infra & data

Apache KafkaRedisElasticsearchPostgreSQLMongoDBDockerKubernetesAWS

the ai bits

LangChainLangGraphRAGVector EmbeddingsAgentic AIPrompt Engineering

practising the craft

JUnit5MockitoDebuggingCode ReviewsAgile / ScrumCI/CD

classroom, recently

Data StructuresAlgorithmsOOPDBMSOperating Systems

chapter five

Sign the guestbook

The internet used to have guestbooks. I’m bringing it back, kind of. If you read this far, drop me a line — a job lead, a critique of my CSS, a project idea, or just a hello. I read every message and try to reply within 48 hours, between sips of chai.

✉   write to rahul →

or email: rparida1052@gmail.com

URGENT

Elsewhere on the web