Open to opportunities

Aakriti Poudel

@aakritipoudel_

Message

AI Engineer building agentic AI workflows and production-ready RAG systems with measurable model improvements.

Nepal

Message

What I'm looking for

I’m looking for a team where I can build production-grade AI—agentic RAG, multi-agent orchestration, and reliable ML pipelines—using strong MLOps (MLflow, DVC, Docker, AWS) and shipping measurable improvements in real workflows.

I’m an AI Engineer with 1+ years of experience specializing in Agentic AI workflows and predictive modeling. I build end-to-end RAG systems, multi-agent orchestration, and fine-tuning approaches that help automate complex processes—backed by LLM output validation and practical reliability techniques like semantic caching.

I’ve shipped production systems that translate business problems into scalable AI: I improved YOLO-based object detection by 10–15% mAP, reduced false positives by 30% with ROI-based OpenCV pipelines, and cut hallucination rates by 40% using an Agentic RAG design with LangChain orchestration, ChromaDB, and Pydantic-validated outputs. I’m also strong in EDA and classical ML (Random Forests, XGBoost), and I apply MLOps best practices with MLflow, DVC, Docker, and AWS to keep workflows clean, testable, and production-grade.

Experience

Work history, roles, and key accomplishments

ML Engineer

Kyra Works Pvt Ltd

Oct 2025 - Mar 2026 (5 months)

Trained and optimized YOLO-based object detection models, improving mAP by 10–15% via dataset balancing, augmentation, and hyperparameter tuning tracked in MLflow. Built ROI-based OpenCV inference pipelines that reduced false positives by 30% on live video streams and supported AWS back-end infrastructure with ML workflows using DVC/MLflow and Terraform.

YOLO Object Detection OpenCV MLFlow DVC Hyperparameter Tuning Terraform

AI Engineer

Freelance

Dec 2024 - Aug 2025 (8 months)

Designed and shipped a production Agentic RAG system using LangChain, ChromaDB, and Pydantic-validated LLM outputs, reducing hallucination rates by 40% and automating end-to-end document workflows. Built structured-data ML pipelines (Random Forest, XGBoost) with MLOps practices using MLflow, DVC, and Docker, and deployed models via FastAPI for client use.

Agentic RAG Lang Chain ChromaDB Pydantic MLFlow DVC Docker fastAPI Random Forest

Education

Degrees, certifications, and relevant coursework

Kathmandu University

Bachelor of Engineering, Computer Engineering

2021 - 2026

Grade: GPA 3.9 / 4.0

Activities and societies: KU HackFest 2023 Education Category Winner (real-time ASL gesture recognition, YOLOv5). Microsoft Learn Student Ambassador (2023–2025) and KU HackFest organising committee (2022, 2024).

BE in Computer Engineering at Kathmandu University (2021–2026) with a GPA of 3.9/4.0. Relevant coursework included AI, machine learning, NLP, computer vision, DBMS, algorithms and data structures, probability and statistics, and software engineering.

DataCamp

Associate Data Scientist, Machine Learning

Completed DataCamp’s Associate Data Scientist program focused on supervised machine learning.

Coursera

Supervised Machine Learning, Machine Learning

Completed the Supervised Machine Learning course by Andrew Ng on Coursera.

Stanford University

Intro to Statistics, Statistics

Completed Intro to Statistics (hosted via Coursera) from Stanford University.

Nepal Applied Mathematics and Informatics Institute for Research (NAAMII)

6th Nepal Annual AI School (ANAIS), Artificial Intelligence

Awarded a full scholarship and selected as a participant in the 6th Nepal Annual AI School (ANAIS), an 11-day intensive program. Covered topics included LLMs and agents, geometric deep learning, and AI for healthcare.