Ram Pujan Barhi

AI & Software Engineer / Python · NLP · Deep Learning

Graduate AI engineer building intelligent systems — LLM pipelines, retrieval-augmented generation and multi-agent workflows — and the production platforms that ship them.

(01)AboutThe human behind the model

The human behind the model

Portrait of Ram Pujan Barhi
Fig. 1 — Ram Pujan BarhiHover to developTap to toggle

Graduate AI engineer passionate about building intelligent systems — with the engineering foundation to bring theoretical depth and practical execution to complex data challenges.

I work across Python, LangChain, retrieval-augmented generation (RAG), LLM integration, and workflow automation. My work spans generative NLP models, sentiment classification, multi-agent frameworks, and data visualization dashboards.

I completed my M.Sc. in Computer Science at UET Lahore with a CGPA of 3.925/4.0, following a B.Sc. in Electrical Engineering. Alongside software engineering, I have dedicated nearly four years to teaching Python, C, data structures, operating systems, and machine learning.

“Engineering the future with AI, one neural network at a time.”

3.925/4.0
CGPA — M.Sc. Computer Science, UET Lahore
04yrs
Teaching Python, C, DSA, operating systems & ML
02live
Client platforms running in production
B.Sc.EE
Electrical Engineering foundation
(02)WorkFeatured work & client highlights

Selected work

Production-grade client platforms and intelligent systems built for real-world impact, high conversion, and scale.

01 Client project / Commercial web platform Live in production

Style Spell Beauty Salon & Academy

Full-stack digital storefront & automated 1-tap WhatsApp booking platform for a premier salon in Lalitpur.

Engineered and launched a high-converting digital storefront and interactive booking architecture for a luxury beauty salon & academy. Replaced manual phone booking friction with real-time multi-category service menus, transparent pricing rate cards, instant 1-tap WhatsApp appointment routing, a dynamic lookbook gallery & nail art studio showcase, student promotional discount flows, and official partner showcases with luxury brands (Kanpeki Japan, Casmara Spain, Schwarzkopf).

Visit live website
stylespellbeautysalon.com.np Style Spell Beauty Salon & Academy production website
  • A

    5.0★ rating with 20+ verified client reviews

  • B

    1-tap WhatsApp instant booking with pre-filled service payloads

  • C

    Interactive menus with tiered pricing for facials, hair & nails

  • D

    Brand authority integration with Kanpeki, Casmara & Schwarzkopf

  • Next.js
  • React
  • Tailwind CSS
  • WhatsApp API
  • UI/UX Design
  • Vercel
02 EdTech platform / Licensing exam engine Live in production

Pharmacist Prep (MCQ Masters)

Adaptive spaced-repetition revision & authentic mock exam engine for the Nepal Pharmacy Council Licensing Examination.

Architected and deployed a comprehensive medical exam preparation platform built strictly according to the official Nepal Pharmacy Council (NPC) 99-MCQ blueprint, for B.Pharm, D.Pharm, and Allied Health licensing candidates. Built with an automated spaced-repetition algorithm (1d → 3d → 7d → 14d → 30d retention intervals), a 90-minute timed mock test simulator with negative marking options, Web Speech API clinical audio synthesis for hands-free study, an automated Mistake Vault, and daily countdown pacing metrics across 8 weighted subject domains.

Visit live platform
mcqmasters.pages.dev Pharmacist Prep MCQ Masters exam simulator
  • A

    99-MCQ mock simulator with 90-min countdown & status palette

  • B

    Spaced repetition (SRS) automated interval decay memory progression

  • C

    Clinical voice narration via the native browser Web Speech API

  • D

    8 council domains with verified pharmacological rationales

  • Next.js
  • TypeScript
  • Tailwind CSS
  • Spaced Repetition SRS
  • Web Speech API
  • Cloudflare Pages

Research & core algorithms

AI & Data Science labs

Explorations across LLM orchestration, agentic pathfinding, and market analytics pipelines.

  • L/01

    NepSentAI & Recipe Crave AI

    Generative-AI-assisted sentiment classification engine for social media text, along with an LLM-powered recipe recommendation system tailored for smart dietary preferences.

    • Python
    • LangChain
    • Streamlit
    • OpenAI
    • Transformers
    AI / NLP
  • L/02

    Butter Bot Explorer & Multi-Agent System

    AI agent pathfinding simulation built with graph search algorithms, alongside a multi-agent orchestration architecture providing operational visibility and automated decision loops.

    • Python
    • A* Algorithm
    • Pygame
    • Agentic AI
    • Dashboards
    Agentic AI
  • L/03

    London Airbnb Market Analysis & RAG

    Interactive Power BI and Python analytics dashboard analyzing rental pricing trends, geographic clusters, and demand forecasting, with integrated retrieval-augmented generation pipelines.

    • Power BI
    • Python
    • Pandas
    • Data Viz
    • LangChain
    Data Science
(03)ExperienceEngineering, research & data analytics

Career path

A track record of engineering intelligent solutions, research, and data analytics.

  1. Current

    AI & Software Engineer

    Independent projects & client solutions

    • Engineers and deploys live client production web platforms including Style Spell Salon (e-commerce & WhatsApp booking) and Pharmacist Prep (spaced-repetition mock exam engine).
    • Architects multi-agent decision pipelines, LLM/RAG workflows, and interactive monitoring dashboards.
    • Translates raw business data into high-accuracy predictive ML models and modern responsive web systems.
    • Next.js
    • React
    • Python
    • LangChain
    • RAG
    • Cloudflare

  2. Mar 2024 — Jun 2024

    Data Analyst

    University of West London

    • Analyzed complex datasets to extract actionable operational insights and trend models.
    • Developed automated visualization dashboards using Power BI and Python.
    • Collaborated with academic researchers on data pipelines and statistical validation.
    • Power BI
    • Python
    • SQL
    • Data Viz

  3. Sep 2021 — Jan 2022

    Data Analyst Intern

    Protrainy

    • Cleaned, processed, and evaluated multi-source relational databases for analytics reporting.
    • Constructed exploratory data analysis (EDA) scripts using Pandas and Seaborn.
    • Optimized SQL queries, improving reporting query speeds by 30%.
    • Python
    • Pandas
    • SQL
    • EDA

  4. 4 years

    Teaching Assistant & CS Tutor

    UET Lahore & independent instruction

    • Taught Python, C, data structures & algorithms, operating systems, and ML.
    • Mentored students through hands-on coding assignments and algorithm analysis.
    • Python
    • C
    • DSA
    • Machine Learning
(04)ToolkitLanguages, frameworks, data tools & platforms

The toolkit

AProgramming languages

  • Python
  • SQL
  • JavaScript
  • R

BAI / ML frameworks

  • TensorFlow
  • PyTorch
  • Scikit-learn
  • Hugging Face & NLP

CData & visualization

  • Power BI
  • Pandas & NumPy
  • Plotly / Seaborn
  • Data pipelines & ETL

DTools & platforms

  • Git & GitHub
  • LangChain & Vector DBs
  • Streamlit & APIs
  • LaTeX & academic research
(05)ContactOpen to conversations

Let’s build something intelligent.

Write to me

contact@rampujanbarhi.com
Say hello