Initializing engineering sandbox...

<B/S> Menu
Download Resume
Ready for Work / Internships

Hi, I'm Bishal Somare Magar

$> |

Currently pursuing a Bachelor in Information Management in Nepal, building production-grade Python backends & integrating RAG-orchestrated Generative AI workflows with LangChain and FastAPI.

View Projects Download Resume
interpreter.py
01 import engine 02 from lang_chain import RAGPipeline, AgentCoordinator 03 04 class DeveloperProfile: 05 name = "Bishal Somare Magar" 06 focus = ["Backend Architectures", "AI Workflows"] 07 degree = "Bachelor of Information Management (BIM)" 08 location = "Kathmandu, Nepal" 09 10 def get_capabilities(self): 11 return RAGPipeline( 12 orchestrator="LangChain", 13 frameworks=["FastAPI", "Django"], 14 agentic=True 15 ) 16 17 # System status verification 18 bishal = DeveloperProfile() 19 print(f"Model active. Local: {bishal.location}")
>>> Executing profile self-check... SUCCESS
FastAPI / Django
LangChain Agentic
/assets/images/profile.jpg
<about_me>

Engineering Mindset

Architecting Bridges Between Data and Intelligence

I am currently studying for my Bachelor of Information Management (BIM) at Tribhuvan University. My engineering journey is driven by a profound curiosity for backend architectures and dynamic intelligence integrations. I build robust Python servers and orchestrate modern Retrieval Augmented Generation (RAG) pipelines that securely process structures into smart decisions.

I avoid simple templates and "starter kits", opting to build systems from the core. I integrate vector storage, manage semantic loaders, orchestrate agent pathways, and deploy scalable REST endpoints. I enjoy diving into deep workflows, managing memory caching with Redis, designing PostgreSQL schemas, and exploring multi-agent architectures using LangGraph.

Current Objective & Research Focus
  • RAG Enhancements: Multi-query retrieval, hybrid sparse-dense searching, and reranking workflows.
  • Agentic Graphs: Building state-aware routing, execution loop, and self-corrective coding agents.
  • Scalable Backend: High-concurrency async operations using FastAPI, Celery scheduling, and Redis.

Academic & Learning Journey

2021 – 2026 Expected
Bachelor of Information Management (BIM)

Tribhuvan University, Nepal

Merging business technology with developer engineering. Hands-on foundations in software development, databases, advanced programming, data models, operations research, and enterprise information architectures.

2019 – 2021
Higher Secondary Education (+2 Science)

NEB Tillottama Higher Boarding School, Butwal

Focusing on physics, mathematics, and computing sciences. Built initial algorithmic foundations and analytical methodologies.

2023 - Present
Specialized Self-Directed Research

Global Open Source & AI Ecosystem

Deep-diving into Generative AI workflows, LangChain, vector storage, and stateful multi-agent graphs to build credible agent platforms.

<skills_inventory>

Modern Technology Stack

Pragmatic software delivery requires robust language knowledge. Organized by architecture segments without bloated rating scales.

GenAI / LLM Systems

  • Generative AI
  • LangChain Framework
  • RAG Pipelines
  • LangGraph Routing
  • Vector DBs / Embeds
  • Agentic Orchestration
  • Prompt Engineering

Backend Architectures

  • Python (Core)
  • Django Framework
  • FastAPI / REST APIs
  • PostgreSQL / MySQL
  • Redis In-Memory
  • API Authentication
  • Backend Caching

Infrastructure & Cloud

  • Docker Container
  • GitHub Actions
  • Docker Compose
  • GitHub Pages
  • Basic Cloud VM
  • Hosting Ecosystem
  • App Configuration

Web Frontend

  • HTML5 Semantic
  • CSS3 Custom Model
  • JavaScript (ES6+)
  • DOM API Events
  • Flexbox & Grid Layout
  • Responsive Views
  • JSON Data Exchange

Engineering Workspaces

  • Git Version Control
  • GitHub Repositories
  • Linux Terminal / Bash
  • VS Code Workspace
  • Postman API Testing
  • Markdown Docs
  • Environment Variables
<production_ready_code>

Showcase Portfolio & Case Studies

Detailed deep-dives into actual implementations. No fake starter configurations—engineered systems built with state structures, problems resolved, and clean outcomes.

<RAG_PIPELINE />
CHROMA DB
Python LangChain FastAPI VectorDB

LLM-Powered Document Retrieval System

A high-accuracy Retrieval-Augmented Generation (RAG) orchestration pipeline resolving static document parsing and intelligent semantic querying.

Problem: Extracting contextual accurate answers from multi-page PDFs without hallucination.
Solution: Split text pipelines with recursive character chunking, embeddings vectors, and optimized search retriever context.
URL
<DJANGO_MANAGEMENT />
POSTGRESQL
Python Django RelationalDB Autosend

Sales & Customer Management Platform

Enterprise-focused CRM for unified inventory tracking, customer accounts lifecycle, vendor coordination, and sales invoice workflows.

Problem: Fragmented tracking of client transactions, inventory stocks, and vendor invoice notifications.
Solution: Robust database relational schema models joined with automated async SMTP channels for low-stock triggers.
URL
<AI_CODE_ASSISTANT />
LLM API
Python Django REST APIs Assistant API

AI System Code Assistant

Integrated developer workflows module mapping technical inquiries to custom generated workspace modules, automated analysis, and diagnostics.

Problem: High context-switching overhead reading complex repository structures, explaining core configurations.
Solution: LLM prompt integration schemas loading active folder trees, offering contextual codebase refactoring structures.
URL
<PREMIUM_ML />
FASTAPI + ML
FastAPI Python ML Model Streamlit

Insurance Premium Predictor

A productionized ML forecasting pipeline mapping numerical client profiles into live computed premium budgets served via a fast microservice structure.

Problem: High manual delays evaluating underwriting hazards, dynamic medical factors, and risk pricing.
Solution: Pre-trained regression estimators exposed through unified API contracts and interactive Streamlit panels.
URL
<MULTI_AGENT_SIM />
LAYOUT TEMPLATE
LangGraph Multi-Agent StateMachines WebSocket

[Interactive Template] Multi-Agent Graph

High-end design showcase model for stateful LangGraph node execution routers, sub-goal generation, and multi-agent loops.

Problem: Custom placeholders showing layout structures for complex state machine execution models.
Solution: Multi-role loops routing verification prompts across independent testing agents.
Ready for Upload
<REDIS_GATEWAY />
LAYOUT TEMPLATE
Python Redis RateLimiter AuthToken

[Interactive Template] High Performance Gateway

Designed template block verifying rate-limiting algorithms, Redis sliding window cache models, and low-latency validation gateways.

Problem: Preventing API endpoint query starvation, keeping responses under microsecond ranges.
Solution: Redis-cached access credentials matching clients using sliding-window rate limit counters.
Ready for Upload
<lang_graph_routing>

Interactive Multi-Agent Workflows

Demonstrating dynamic decision loops in multi-agent routing systems using state charts. Interact with the nodes below to trigger the compiler simulation.

1. User Query
2. Coordinator / Router
3. Semantic Search
4. Guardrail Grader
State Engine Execution Logs
[09:18:47] INITIALIZED: StateCoordinator active. [09:18:48] Info: Ready to route test prompt indices. Click nodes to run single-step verification.
<milestones_and_growth>

Training & Certifications Timeline

Documented professional workshops, learning achievements, and academic timelines verifying technical commitment.

Education 2021 – 2026

Bachelor of Information Management (BIM)

Tribhuvan University, Nepal

  • Comprehensive education in algorithms, systems analysis, data structures, and computer architectures.
  • Applied academic concepts by building Django systems and relational database operations.
Certification Placeholder 2024

DeepLearning.ai / LangChain Certifications

Specialized LLM & RAG Training Units

  • Verified capability designing prompt-templates, embeddings search optimization, and evaluation scoring models.
  • Certificate Image: /assets/resume/cert1.jpg
Workshops / Training 2023 - Present

Backend & DevOps Specialization Workshop

Distributed Systems, SQL, and Docker

  • Hands-on training building containerized test models, mapping network boundaries, and caching database views with Redis.
  • Practiced PostgreSQL tuning, REST API versioning, and secure user sessions logic.
Secondary Years 2019 – 2021

NEB Tillottama Higher Boarding School

Butwal, Nepal (+2 Science)

  • Broad education covering mathematics, physics, and programming foundations.
  • Graduated with solid marks, preparing for higher university systems engineering.
<repository_analytics>

GitHub Engagement & Contribution Activity

Live-inspired matrix representation of coding habits, version control updates, and repository delivery metrics.

Contribution Calendar Simulation 324 Contributions / Year
Less Active
Highly Active
Commits (YTD)
1,480+
PR Accept Rate
98.4%
Repo Stars & Clones
48+
Inspect GitHub Repositories
<get_in_touch>

Establish Collaboration

Have an open internship, a remote software backend project, or want to discuss LLM orchestration pipelines? File a transmission below.

Direct Connection Coordinates

I am located in Kathmandu and open to remote work structures, junior backend/AI roles, and active project collaborations. Feel free to reach out via these direct channels or drop a message through the workspace client on the right.

Email Transmission bishalsomare09@gmail.com
Phone Telemetry +977 9848134815
Geographic Center Kathmandu, Nepal
$> ping -c 3 bishal.somare > Success [0ms]