Skip to content
Botmer International®
Botmer talent profile

Junaid Rafique

Senior AI & Machine Learning Engineer

AI and machine-learning engineer with 8+ years across recommendation systems, NLP, computer vision, healthcare, fintech, and SaaS. Builds data pipelines, vector search, conversational assistants, and production APIs. Experience includes FastAPI, Qdrant, PostgreSQL, LLaMA, knowledge graphs, cloud deployment, and model optimization with MLOps tooling.

8+ years
AI and machine-learning experience
Recommendations & RAG
retrieval and conversational systems
FastAPI & Qdrant
production service and search stack
Meet the engineer

AI engineering with production depth

Where Junaid adds value

  • Builds book recommendation infrastructure combining metadata ingestion, embedding pipelines, Qdrant similarity search, PostgreSQL, live interaction data, and behavioral user modeling.
  • Develops conversational assistants with LLaMA and Mistral, integrating vector retrieval and natural-language interactions into production recommendation and knowledge workflows.
  • Designs ETL, data schemas, FastAPI services, unit tests, and K6 performance tests, with Docker delivery and cloud-oriented MLOps practices.
  • Applies NLP, knowledge graphs, predictive modeling, and anomaly detection to transport optimization, commodity analytics, document question answering, and healthcare data systems.
Explore professional experience
Core competencies

AI systems grounded in software engineering

01

Recommendation and vector search

Builds content-based and collaborative filtering systems with embeddings, metadata pipelines, behavioral signals, PostgreSQL, and Qdrant to support real-time personalization and contextual retrieval across book discovery workflows.

  • Qdrant
  • PostgreSQL
  • Embeddings
  • Collaborative Filtering
  • FastAPI
  • ETL
02

LLM and knowledge applications

Develops conversational assistants, RAG applications, document question answering, and natural-language knowledge-graph queries using LLaMA, Mistral, LangChain, OpenAI APIs, and retrieval integrations for business users.

  • LLaMA
  • Mistral
  • LangChain
  • RAG
  • Knowledge Graphs
  • OpenAI API
03

Machine learning and vision

Applies predictive models, feature engineering, hyperparameter tuning, anomaly detection, NLP, and computer vision using Python, Scikit-learn, TensorFlow, PyTorch, XGBoost, and OpenCV across varied analytical domains.

  • Scikit-learn
  • TensorFlow
  • PyTorch
  • XGBoost
  • OpenCV
  • NLP
04

Production ML engineering

Builds data ingestion, API layers, containerized deployment, automated tests, load testing, and model-serving workflows, supported by cloud platforms, Kubernetes, CI/CD, and model monitoring tools in production.

  • Docker
  • Kubernetes
  • K6
  • MLflow
  • FastAPI
  • CI/CD
Experience

From models to product delivery

01/2024 – presentCachelogic

Senior Machine Learning Engineer

Leads book recommendation and conversational AI development, building data infrastructure, vector pipelines, personalized models, tested FastAPI services, and containerized deployments while contributing Slack and Jira integration to internal RAG workflows.

  • Ingested and normalized book metadata from Amazon, Ingram, and Goodreads using ETL and Alembic schemas.
  • Built Qdrant embeddings and real-time content-based and collaborative recommendations.
  • Shipped Nora AI using LLaMA 3 8B on A6000 GPUs.
  • Built unit and K6 performance tests and deployed Docker services to private cloud.
05/2020 – 12/2023Devbatch

Senior Data Scientist

Led AI projects for international clients across transport, agriculture, document retrieval, and commodities analytics, combining predictive models, knowledge graphs, RAG, anomaly detection, stakeholder communication, and Python-based analytical workflows.

  • Analyzed 5GB daily AVL data for Hajj bus operation optimization.
  • Designed extractive Arabic document question answering with LangChain RAG and open-source LLMs.
  • Built farmer-churn prediction with XGBoost, SVM, and logistic regression.
  • Developed a ChatGPT-based NLQA system integrated with knowledge graphs for commodity data.
02/2017 – 04/2020PureLogics

AI Engineer

Developed AI and machine-learning models and end-to-end pipelines, including preprocessing, feature engineering, training, evaluation, inference optimization, ETL, cloud deployment, technical documentation, and collaboration with product and engineering stakeholders.

  • Designed and deployed ML pipelines from data preparation to evaluation and integration.
  • Worked on NLP, computer vision, and deep-learning solutions using TensorFlow and PyTorch.
  • Supported model optimization, production monitoring, code review, and junior engineer mentoring.
Selected work

AI systems built for practical use

Highlights from professional projects, product delivery, and specialist work.

Conversational recommendations

Nora AI

Developed a book assistant using LLaMA 3 and Mistral 7B, with embeddings and real-time filtering. Integrated Qdrant, PostgreSQL, and FastAPI to deliver personalized recommendations and natural-language interactions linked to user goals and preferences.

  • LLaMA 3
  • Mistral 7B
  • Qdrant
  • FastAPI
Recommendation infrastructure

Real-time Personalized Book Recommendation Engine

Built metadata ingestion and embedding workflows for book discovery, combining content-based and collaborative filtering with live engagement data. Delivered tested and load-tested APIs with hybrid search techniques and containerized deployment in a private cloud environment.

  • PostgreSQL
  • Qdrant
  • Embeddings
  • Docker
Internal knowledge automation

Organizational Bot for Slack and Jira

Built an assistant to query Slack and Jira for ticket status, workload information, and channel analytics. Connected PostgreSQL and FastAPI with the respective APIs, supporting stakeholder visibility into operational work and internal communication data.

  • FastAPI
  • PostgreSQL
  • Slack API
  • Jira API
NLP and retrieval

Arabic NLP RAG

Designed an extractive question-answering system for Arabic religious texts using LangChain retrieval-augmented generation and open-source language models. The implementation focused on document retrieval and improved access to relevant textual information for Arabic-language question answering.

  • LangChain
  • RAG
  • NLP
  • Open-source LLMs
Education & professional focus

A foundation for practical delivery

Education & qualifications

Master's in Computer Sciences

Comsats University

08/2018 – 08/2020

Bachelor of Computer Science

Comsats University

04/2014 – 06/2018

Professional focusPersonalized retrieval and conversational systems

Connects embeddings, vector databases, metadata ingestion, and behavioral modeling with LLM interfaces to support book discovery, personalized recommendations, and contextual conversations in production services.

Professional focusEnd-to-end machine learning delivery

Works from data ingestion and feature engineering through API development, tests, load validation, and deployment, with applied experience across transport, healthcare, agriculture, and SaaS.

Talk to Botmer

See if Junaid fits your AI engineering team

Share your requirements, current challenges, and delivery goals. Botmer will review the brief and help arrange the next step.

Request a talent match