RJRajan Jasani
Backend systemsApplied AIPune / India

ENGINEERINGSYSTEMS THATTHINK, RETRIEVEAND SCALE.

I design backend and AI systems where retrieval, orchestration, data and infrastructure have to work as one production system.

Selected work Résumé

Intelligence

Retrieval, model calls, reranking and structured outputs.RAG · Bedrock · Embeddings

What I do

Open the system. See how it works. My portfolio is built around architecture, technical decisions and the constraints that shaped the final product.

Selected systems

Three kinds of proof: production work, private client systems and technical experiments.

Production knowledge system2025—26

Ask HipHop

A citation-first research system for a domain where context, source quality and interpretation matter as much as retrieval speed.

I worked across API design, asynchronous ingestion, multi-knowledge-base retrieval, metadata filtering, reranking, source provenance and grounded response generation.

60+textbooks
500+hours video
40+backend APIs
Read engineering case study
chat.ask.hiphopproduction
Ask HipHop product interface

Request path

Trace one answer through the system.

Question
Classify
Plan
Retrieve
Rerank
Generate
Cite
Private client system

BinaAI

BinaAI interface preview

Curriculum-aware AI exam preparation with grounded answers, citations, quizzes and controlled access.

The public case study explains the architecture and development process without publishing private code, client data or confidential prompts.

Problem
Ground answers inside a strict exam curriculum.
System
Django + FastAPI services, PostgreSQL, vector retrieval and React.
Focus
Scope classification, citations, structured outputs and usage gating.
Disclosure
Architecture only; client-sensitive details omitted.
Read sanitized case study
Hybrid RAG Playground

Inspectable vector + lexical retrieval experiments.

Engineering Systems Video Pipeline

Programmatic technical-video production workflow.

Measurement Management System

Field reporting and offline sync for civil engineering operations.

client system

Engineering notes

Writing tied to systems I have actually built, debugged and operated.

Retrieval architecture

Why vector similarity is only the beginning

Classification, retrieval planning, metadata constraints and reranking inside a domain-specific RAG system.

Drafting
Private client systems

How to explain a closed AI product without leaking the client

A technical case-study pattern for architecture, safeguards, evaluation and product behavior without exposing sensitive implementation details.

Drafting
Backend systems

Async ingestion is a product capability, not a background detail

Retries, state transitions, observability and failure recovery for document-heavy AI applications.

Drafting

Backend first.
Product aware.

I work mainly with Python, FastAPI, Django, PostgreSQL, Redis and AWS, with applied AI focused on retrieval, agents and production integration rather than model demos in isolation.

CurrentSoftware Engineer L2 · Crest Infosystems
FocusBackend · Applied AI · System Design
CommunityDjango India · DjangoDay India
CertificationAWS Certified AI Practitioner
PublishingEngineering Systems