2027 EXPECTED
Computer Science & Engineering
Nirma University · Gujarat, India
Building foundations in software, systems, and computing. SOFTWARE ENGINEER · FULL-STACK · APPLIED AI
BASED IN INDIA
Parth Srivastava.
I build products that connect systems and intelligence.
01 / ABOUT & EXPERIENCEEngineer by
Engineer by
curiosity.
The through line is simple:
make the complicated useful.
I'm a computer science student who cares about what happens under the hood.
My work sits between full-stack product development, backend systems, and applied AI. I enjoy understanding how components fit together — and making that complexity useful to the person on the other side of the screen.
Currently studying at Nirma University and learning through building, internships, and the questions that show up along the way.
MORE TO BUILD
TOOLS I REACH FORTHE RIGHT TOOL DEPENDS ON THE PROBLEM
01
LANGUAGES
JavaScript · Python · C++ · Java · SQL
02
FRONTEND & FULL-STACK
React · Next.js · Tailwind CSS · shadcn/ui
03
BACKEND & APIS
Node.js · Express.js · FastAPI · REST · JWT
04
DATA & INFRASTRUCTURE
PostgreSQL · MongoDB · Redis · Docker · AWS · Git · CI/CD
05
AI & MACHINE LEARNING
LLMs · RAG · Embeddings · Vector search · Transformers · PyTorch · LangChain · LangGraph
02 / SELECTED WORK
Selected work.
Systems built to solve real problems.
A look at the thinking behind them.
01 — EMPLOYEE INTELLIGENCE / RAGCASE STUDY / AI SYSTEMS
Sahay
A contextual assistant concept for helping HR teams find useful signals across the tools where work happens.
ENGINEERING FOCUS
Exploring ingestion, retrieval, permission-aware context, and useful LLM summaries across workplace data.
REPOSITORY URL TO CONFIGURELive demo
SAHAY / CONTEXT FLOWILLUSTRATIVE
#Slacksource 01
▦Calendarsource 02
◎Peoplesource 03
◷Meetingssource 04
01INGESTNormalize events
02RETRIEVESemantic + filters
03RESPONDGrounded context
CONTEXT WINDOWPS
grounded summary
PERMISSIONS FILTEREDNO LIVE DATA · DEMO CONCEPT
Project details are based on the supplied brief.
02 — TRAFFIC INTELLIGENCE / BACKENDCASE STUDY / BACKEND
MargDarshak
A traffic co-pilot built around FastAPI, REST endpoints, and congestion-aware route comparisons.
ENGINEERING FOCUS
Backend engineering and API design for a traffic intelligence project described as having approximately 25 REST APIs.
REPOSITORY URL TO CONFIGURELive demo
MARGDARSHak / ROUTE STUDYILLUSTRATIVE MAP
START / AHD-04
DESTINATION
ROUTE COMPARISON DEMO
Route Alower congestion
18 minRoute Bshorter distance
22 minFASTAPI / ROUTE COMPUTATIONNO LIVE TRAFFIC DATA
Route artwork is illustrative, not live map or traffic data.
03 / ENGINEERINGBeyond the
Beyond the
interface.
Good software is a chain of decisions.
Here's what I pay attention to.
ENGINEERING NOTESSELECT A COMPONENT
SAHAY · CONCEPTUAL
Retrieve with useful context
A RAG pipeline combines semantic similarity with metadata filters. The result is a smaller, more relevant context window for an LLM response.
INPUT
RETRIEVAL FLOW
OUTPUT
ENGINEERING IS TRADE-OFFS MADE VISIBLE→
03.1Think in flows.
Think in flows.
Design for failure.
From an API request to an AI-generated answer, each boundary changes what the system can promise. I like making those boundaries legible.
RETRIEVAL · API DESIGN · DATA FLOW04 / OPEN CHANNEL AVAILABLE FOR GOOD CONVERSATIONS
Have a problem
worth building?
Tell me what you're figuring out.
I'm always up for talking systems, products, and ideas.
ON EMAIL