Portfolio · 2026
RajdeepHundal
Full Stack Developer · HCI
Multimedia -> Full Stack -> HCI
Bidge full stack development and human-centred design — because the best technology should feel invisible.
Available Immediately
"Nature never designs anything
without purpose." — Inspired by Aristotle
RH
§1 — Axioms
Given the following
These are the foundational tools and languages. Everything else in this portfolio is derived from them.
AXIOM 1.3
Flash Projects
View Flash (older) Projects
O(log n)
AXIOM 1.4
AngularJS
View AngularJS With Java Backend ~ Library Directory Application for Texas A&M University.
Library Directory Application
∂L/∂w → 0
AXIOM 1.5
Full Stack
View Full Stack Projects with Java backend and React Frontend.
View Projects
Av = λv
AXIOM 1.6
Floral Design
Teaches composition & negative space ~ currently pursuing a career diploma in Floral Design at Penn Foster.
∑ petals = φⁿ
Experience
2025 — 2026
FDM Group
IT Consultant (Software Developer and QA Automation)
2023 — 2024
Texas A&M University
Software Applications Developer I
Contributed to in-house and open-source applications used by students, faculty, and internal teams such as sprint calculator, library directory, FOLIO and other applications.
2023
Cook Systems
Full Stack Apprentice
2022
HCl America Inc.
Project Lead
Led and coordinated IT infrastructure deployment initiatives for Ascension hospitals and outpatient clinics, supporting mission-critical healthcare environments via HCL.
Education
The University of Winnipeg - PACE
Post Graduate Diploma in Project Management and Certificate in Management
2018 — 2019
Relevant coursework: Project Management, Organizational Behaviour, Leadership, Business Communications.
View Program
University of Wisconsin - Superior
BSc, Computer Science
2004 — 2007
Collaborated in a team to design and develop a full-stack Insurance Calculator for a fictitious client as a capstone. The app calculated premiums, and compared insurance plans — built with HTML/CSS/JS, C# .NET, and SQL.
Selected Projects
01
Twitter Clone
02
Project Management Application
03
Open Source - FOLIO
Contributed to local as well as community sprints. I contributed to FOLIO while at Texas A&M — it's an open-source platform used by research libraries worldwide, so the work had to meet community standards, in addition to internal ones.
View Projects FOLIO Contributions
Certifications & What I'm Currently Learning
2025 · ByteByteGo
AI Engineering Bootcamp
Building practical AI systems — from LLM foundations and RAG pipelines to reasoning agents and multi-modal generation
5
Projects Built
6
Weeks Intensive
01
LLM Playground
LLM Foundations · Tokenization · Transformer Architecture
↓
Tokenization
Word-level tokenization
Character-level tokenization
Subword-level tokenization
BPE with TikToken
Vocabulary building
Model Architecture & Generation
Linear & Transformer layers
GPT-2 internals
Loading pretrained models via HuggingFace
Greedy decoding
Top-k & top-p sampling
Base vs instruction-tuned models (Qwen3 vs GPT-2)
RAG Pipeline
Document ingestion (PDF, URL, text)
Text chunking strategies
Embedding generation
FAISS vector index
Retrieval-Augmented Generation chain
Generation & UI
Ollama local model serving (Gemma3-1B)
LangChain ConversationalRetrievalChain
System prompt engineering
RAG chain validation
Streamlit web UI
Tool Calling
Why tool calling matters
Implementing custom tools
Function schemas via docstrings & type hints
Parsing LLM tool call output
LangChain tool integration
Agentic Systems
Agents vs agentic systems vs LLMs
Agency levels
Multi-step agent workflows
Perplexity-style web search agent
Minimal backend & frontend UI
04
"Deep Research" Capability
Reasoning Models · Inference-Time Scaling · Web Search
↓
Inference-Time Scaling
Few-shot & zero-shot CoT prompting
Self-consistency sampling
Sequential revision
Tree of Thoughts (ToT)
CoT on non-reasoning models (GPT-2)
Reasoning Models & Training
DeepSeek-R1 & OpenAI "o" family overview
STaR (Self-Taught Reasoner)
ORM vs PRM + RL
Deep research agent with web search
Multi-agent deep research (optional)
05
Multi-Modal Generation Agent
Text-to-Image · Text-to-Video · Routing Agent · Gradio
↓
Image & Video Generation
Stable Diffusion XL (SDXL)
Text-to-image generation & inference steps
Visualizing the diffusion process
Text-to-video (damo-vilab model)
Frame inspection & MP4 conversion
Multi-Modal Agent
Intent-based request routing
LLM for generic queries
Unified multi-modal agent
Gradio UI
HuggingFace Diffusers integration
OMSCS Course Journey
Spring2027
Semester 1 — Foundations
Data Analytics3cr
TBD3cr
Fall2027
Semester 2 — Specialization
TBD3cr
TBD3cr
Spring2028
Semester 3 — Specialization
TBD3cr
TBD3cr
Fall2028
Semester 4 — Specialization
TBD3cr
TBD3cr
Spring2029
Semester 5 — Capstone
TBD6cr
Degree Progress
0 / 30 cr