Jasmanjot Singh Sarna
ResumeGet in Touch
Final-Year B.Tech CSE (AI & ML), 2023–2027•Jaipur, Rajasthan, India

Jasmanjot

Singh Sarna
Jasmanjot Singh Sarna
Jasmanjot S. SarnaJaipur, IN
Focus:AI/ML Engineer & Systems Builder

I work across the pipeline: data cleaning and EDA, statistical modeling, LLM/agentic workflows, and full-stack software engineering. I've worked as an AI Full Stack Developer building LLM chatbots and as a Data Analyst Intern building KPI dashboards.

Download Resume
GitHub
LinkedIn
Email
Fast Proof:01. CareerOS (10-Module SaaS)02. Multi-Agent Pipeline
Python•
LangChain•
Next.js 15•
NestJS•
PyTorch•
TensorFlow•
OpenAI API•
PostgreSQL•
Redis•
BullMQ•
Scikit-Learn•
Pandas•
OpenCV•
SQL•
Power BI•
Python•
LangChain•
Next.js 15•
NestJS•
PyTorch•
TensorFlow•
OpenAI API•
PostgreSQL•
Redis•
BullMQ•
Scikit-Learn•
Pandas•
OpenCV•
SQL•
Power BI•
Python•
LangChain•
Next.js 15•
NestJS•
PyTorch•
TensorFlow•
OpenAI API•
PostgreSQL•
Redis•
BullMQ•
Scikit-Learn•
Pandas•
OpenCV•
SQL•
Power BI•
Projects

Top Projects

Full-stack AI SaaS platforms, multi-agent systems, and production machine learning applications.

Plate 01•Full Stack & AI Developer
May 2026 – PresentRole: Full Stack & AI DeveloperIN DEVELOPMENTFEATURED PLATFORM

CareerOS

AI Career Operating System (10-Module SaaS Platform)

Built both the frontend (React/Next.js 15) and backend (NestJS) for a 10-module SaaS platform, including the AI Resume Suite, Interview Suite, and Job Tracker modules. Integrated the OpenAI API with prompt engineering to power an AI ATS Resume Builder module.

How it works // Simplified Overview (In Development)Simplified Overview
In Development
← Swipe to view full flow →
Frontend
React / Next.js 15
API requests
Backend API
NestJS
Background jobs
Async Queue
BullMQ + Redis
Prompt payloads
AI Engine
OpenAI API
Structured response
Database
PostgreSQL
To confirm:Currently in active development with backend queue and database schema wiring.
ReactNext.js 15NestJSPostgreSQLRedisBullMQOpenAI API
In Development
Plate 02•AI Engineer
2026Role: AI EngineerIN DEVELOPMENT

Multi AI Agent Research System

Autonomous Multi-Agent Research Pipeline

Built a multi-agent research pipeline using LangChain and GPT-4o-mini, with search and reader agents collecting data and writer and critic agents producing the final report.

How it works // Autonomous Agent Topology (In Development)Simplified Overview
In Development
← Swipe to view full flow →
Search Agent
Data collection
Raw sources
Reader Agent
Data collection
Parsed data
Writer Agent
Report production
Draft output
Critic Agent
Report production
To confirm:Agent message-passing protocol and tool-calling loop in active development.
PythonLangChainGPT-4o-miniVector DBMulti-Agent
In Development
Plate 03•ML & Full Stack Developer
2026Role: ML & Full Stack Developer

Cardio Detect AI

Clinical Intelligence & Cardiovascular Risk Classification Platform

Engineered a medical-technology web platform exploring cardiovascular disease risk classification through radical algorithmic interpretability. Trained a reproducible Scikit-Learn K-Nearest Neighbors pipeline (k=5, AUC = 0.9269, 86.41% accuracy) over historical clinical cohorts (918 cases), served via FastAPI. Developed a high-precision React 19 / Vite frontend with real-time canvas ECG telemetry, 5 nearest neighbor consensus, and an interactive What-If sensitivity simulator.

How it works // Clinical ML Inference FlowSimplified Overview
Production Architecture
← Swipe to view full flow →
Patient Vitals
11 Clinical Biomarkers
Telemetry POST
FastAPI Backend
Uvicorn ASGI
Normalized feature vector
KNN Pipeline
Scikit-Learn (k=5)
Euclidean nearest neighbors
Interpretability
5 Neighbor Consensus
Risk & sensitivity verdict
Clinical Console
React 19 + Real-time ECG
To confirm:Production model deployed and validated against holdout clinical cohorts.
React 19VitePythonFastAPIScikit-LearnTailwind CSSRecharts
SourceLive
Plate 04•ML & Full Stack Developer
2026Role: ML & Full Stack Developer

Mind Scroll AI

Student Social Media & Mental Health Impact Predictor

Built a full-stack machine learning application trained on 5,000 student survey responses to analyze digital lifestyle impacts on mental health. Trained and evaluated multiple regression architectures, deploying a tuned Random Forest regression pipeline (R² = 0.864, MAE = ±0.37) with ColumnTransformer preprocessing. Engineered a FastAPI backend and React 19 frontend featuring an animated semicircular wellness score gauge, waterfall driver analysis, and real-time What-If habit simulation.

How it works // Predictive Regression PipelineSimplified Overview
Production Architecture
← Swipe to view full flow →
Habit Telemetry
12 Lifestyle Features
Student profile
Preprocessing
ColumnTransformer
Transformed feature matrix
Random Forest
Tuned Regressor (R²=0.864)
Predicted wellness index
Driver Engine
Impact Waterfall
What-if feedback & PDF
Interactive UI
React 19 + Score Gauge
To confirm:Evaluated with 70/30 train/test split on empirical student dataset.
React 19VitePythonFastAPIScikit-LearnTailwind CSSFramer Motion
SourceLive
Proof of Systems Thinking•Multi AI Agent Research System

Interactive Agent Pipeline Explorer

An interactive inspection tool demonstrating how the Multi AI Agent Research System divides complex research into four cooperating LangChain agents. Data packets travel between stages as research progresses from raw retrieval to verified sign-off.

Active Stage Flow // Agent Pipeline
Active Stage // 01

Search Agent

Core Responsibility

Deconstructs user hypotheses into targeted query matrices and crawls authoritative sources.

Takes In

High-level research topic, user constraints, and scope boundaries.

Passes On

Ranked search result URLs, snippet metadata, and candidate documents.

INPUT PAYLOAD[Illustrative Example]
{
  "topic": "Evaluating LLM Agent latency in Next.js / NestJS workflows",
  "domain_whitelist": ["arxiv.org", "github.com", "developer docs"],
  "depth": 2
}
OUTPUT ARTIFACT[Illustrative Example]
{
  "generated_queries": [
    "redis bullmq job queue latency benchmarks",
    "langchain multi-agent sequential pipeline overhead"
  ],
  "sources_retrieved": 12
}
LangChain + Python + GPT-4o-mini OrchestrationUse ← / → keys or swipe on touch to step through agents
Section 03•Professional Background

Work Experience

A verified record of hands-on internships across AI full-stack development, exploratory data analysis, and KPI reporting.

AI Full Stack Developer

Lavelda Coffee Beans

Mar 2026 – Aug 2026•Remote
  • •Developed and deployed LLM-powered AI chatbots using the OpenAI API to automate customer support and product inquiries.
  • •Designed prompt engineering workflows to optimize chatbot accuracy and improve response quality.

Data Analyst Intern

AppGallop Pvt. Ltd.

Feb 2025 – Jan 2026•Remote
  • •Performed data preprocessing and EDA in Python/Pandas to support accurate, reliable reporting.
  • •Built and automated KPI dashboards in Power BI, translating raw data into structured insights.

Summer Intern

Samatrix Consulting Pvt. Ltd.

June 2025 – July 2025•Remote
  • •Cleaned and transformed real-world datasets using statistical techniques, preparing structured data for downstream analysis.
Proficiencies

Technical Stack

Generative AI & LLMs

OpenAI APILangChainRAGMulti-Agent SystemsPrompt Engineering

ML & Deep Learning

PyTorchTensorFlowScikit-LearnCNNsBiGRURegressionModel Evaluation

Full Stack Development

ReactNext.js 15TypeScriptNestJSNode.jsREST APIsPostgreSQLRedis

Data & Analytics

PythonPandasNumPySQLPower BIEDAStatistical Analysis

Tools & Practices

Gitpnpm/TurborepoGoogle Colab
Credentials

Certifications

Meta Full Stack Developer Professional Certificate

Meta•Coursera Online

IBM AI Engineering Certificate

IBM•Coursera Online

IBM Data Scientist Professional Certificate

IBM•Coursera Online

Google Data Analytics Certificate

Google•Coursera Online

All certificates verified via Coursera Online. Official credential links can be verified on LinkedIn or supplied upon interview request.

Communication // Channel 01

Get in Touch

I am open to full-time AI/ML engineering, full-stack systems, and data science roles starting upon graduation in 2027, as well as high-impact remote internships, freelance collaborations, and technical research.

Direct Emailjasmanjotsinghsarna@gmail.com
Verified DossierDownload Master Resume (PDF)
Profiles & Code
GitHub•LinkedIn

Send a Direct Inquiry

Persisted securely to database & delivered to Jasmanjot's primary inbox.

0 / 3000 chars

Direct communication guarantee: your note is sent privately and is never shared.