नेपाली
Pranesh Pyara Shrestha

Er. Pranesh Pyara Shrestha

Electronics, Communication & Information Engineer (NEC reg no: 95385)

Email GitHub Twitter Blog Website YouTube

I like building practical software, AI systems, and infrastructure—from mobile apps to cloud automation—and keeping things minimal and fast. I enjoy shipping, iterating, and learning in public.

Work Experience

Oct 2025–Present
AI Engineer
Singapore Ecommerce App Pte. Ltd. · Full‑time
Singapore · Remote
  • Designed and implemented AI features and automations.
  • Integrated models into product workflows; supported evaluation and iteration.
May 2024–Jul 2025
Software Engineer
Algorithm Squad FZE · Full‑time
Dubai, United Arab Emirates · Remote

Education

2021–2025
Bachelor of Engineering
Institute of Engineering, Pulchowk Campus
Aggregate Percentage: 84%
2018–2021
High School / +2
Dharan Secondary School
CGPA: 3.82
Science track with strong focus on mathematics and physics
2005–2018
Schooling
Nepal Police School, Dharan
GPA: 3.80
Formative years: teamwork, discipline, curiosity

Projects Delivered/Worked On

TheOver.AI

Sports analytics app for bettors using AI tips and real-time player performance tracking.

SabdaSagar

Offline Nepali dictionary with 100K+ words and speech-to-text powered search.

Nepal Election App 2079

Provided real-time electoral data with visualization and candidate information during Nepal's 2079 elections.

MCQ Generation using T5 Transformers (Academic Project)

Built a system using fine-tuned T5-small models for automatic question generation. Trained on SQuAD and RACE datasets for question-answer and distractor generation. Used Hugging Face + PyTorch.

Cloud Resource Allocation with PSO (AI Project)

Simulated dynamic cloud resource allocation using Particle Swarm Optimization. Visualized performance with a Flutter-based frontend.

Speech Recognition & Translation App

Deployed on-device speech-to-text and translation models for Nepali–English. Used TFLite for fast, offline performance.

AutoLead AI (Python + LangChain + n8n)

LLM-based lead qualification agent that classifies and routes inbound messages using LangChain + OpenAI. Triggered automated actions in n8n (Airtable, email updates, CRM sync).

DocResponder (Python + FastAPI + RAG)

An RAG-based AI service that answers support queries from documents. Combined FAISS vector search with LangChain and OpenAI for fast, relevant response generation.

Infrastructure Monitoring and Alerting System

Built monitoring system using Prometheus, Grafana, and Alertmanager with extensive Linux command-line configuration. Deployed using Docker containers and Kubernetes with Helm. Created shell scripts for automated monitoring and alert management.

Smart Incident Router (n8n + Slack + PagerDuty)

Automates alert intake from Prometheus/CloudWatch, deduplicates, and classifies severity with LLM. Enriches with runbook context and routes incidents to on‑call via Slack/PagerDuty, with auto‑escalation and Notion post‑mortem summaries.

Contact