Muhammad Zaeem

AI Engineer & ICPC Gold Medallist specializing in RAG, Automation, and Algorithmic Problem Solving.

Based in: Lahore, PakistanCurrently: Automation Engineer, Joblogic Service Management SoftwareExperience: 3 years

An experienced AI Engineer and Data Science graduate with over 2 years of expertise in building Python-based automation, RAG systems, and machine learning pipelines. He excels at converting complex research ideas into reliable production services, backed by a strong competitive programming background as a Codeforces Expert.

Experience

Automation Engineer

Joblogic Service Management SoftwareworkJul 2026 – Present

Summary

Build and maintain Python automation workflows for scheduling, dispatch and service-state operations, translating complex business rules into reliable production behaviour. Engineer FastAPI services and third-party API integrations with structured validation, transformation, error handling, logging, monitoring and automated tests. Work with operational data across interconnected jobs, sites, users and workflow states, ensuring clean hand-offs between internal services and external systems. Collaborate with product, engineering and business stakeholders to investigate process gaps, prototype solutions and turn requirements into maintainable automation features.

What I did

Implemented reliability features including retries, validation, logging, and performance tuning for production systems.

Results

Automated operational workflows end-to-end to remove repetitive manual steps and improve process reliability.
Reduced human error by automating data checks and information transfers between systems.
Saved several hours of manual work each week per workflow.
Handled thousands of records and events through APIs, background jobs, and integrations.

PythonFastAPIREST APIsLoggingMonitoringAutomated TestingGitHub

Software Engineer

7ValsworkJul 2025 – Jul 2026

Summary

Developed Python-based AI-assisted workflows and data-processing services for a large-scale asset-management platform used by thousands of organisations. Built asynchronous FastAPI/Flask microservices for AI and integration workloads, connecting third-party APIs and handling concurrency, rate limits, validation and production failure cases. Created reusable middleware and decorators for authentication, logging, request validation and operational controls across AI-enabled endpoints. Optimised SQLAlchemy-backed database access and query paths, and strengthened reliability with unit, integration and system testing for data-intensive services.

What I did

Developed systems to achieve faster processing and fewer manual interventions.

Results

Reduced repeated engineering and operational work through backend and data-heavy automation.
Enabled workflows to scale without adding equivalent operational overhead.

PythonFastAPIFlaskSQLAlchemyAsynchronous PythonUnit TestingIntegration TestingSystem TestingGitHub

AI/ML Engineer

Future Connect Training & Recruitment LtdworkSep 2024 – Jun 2025

Designed end-to-end machine-learning workflows in Python using scikit-learn, Pandas and NumPy, covering data cleaning, feature engineering, model training, validation and error analysis. Prototyped NLP and deep-learning approaches for real-world business use cases, evaluating model quality, interpretability and deployment trade-offs. Built repeatable analysis workflows and communicated model behaviour through reports and dashboards so technical findings could support business decisions. Worked with stakeholders to translate loosely defined business problems into practical data and AI experiments, iterating from exploratory analysis to testable solutions.

Pythonscikit-learnPandasNumPyNLPDeep LearningDashboardsFine-tuning

Skills

Technical

Python

Git

Pandas

NumPy

scikit-learn

REST APIs

Vector Retrieval

FastAPI

LangChain

Asynchronous Python

PostgreSQL

Flask

Groq LLM

Prompt Engineering

Embeddings

pgvector

Reranking

NLP

SQLAlchemy

Docker

Postman

Deep Learning

GitHub Actions

Selenium

Jira

Streamlit

MongoDB

Plotly

Matplotlib

Fine-tuning

Model Evaluation

Supervised Learning

System Design

Testing

TypeScript

React Native

Expo

Areas of expertise

Algorithmic Problem Solving

Projects

PredictOps — ML-Driven Predictive Maintenance Dashboard

Lead Developerpersonal

Developed a predictive-maintenance pipeline from historical sensor and maintenance data, including preprocessing, feature engineering and supervised model training. Compared Random Forest and XGBoost models using precision-recall trade-offs and error analysis to prioritise high-risk assets rather than relying only on aggregate accuracy. Served real-time predictions through FastAPI to a dashboard with automated alerts, connecting model outputs with operational decision-making workflows. Added retraining and monitoring components so model behaviour could be reassessed as new operational data arrived and data distributions changed.

PythonFastAPIscikit-learnPandasPostgreSQLReact

BRIEFLY — Transformer-Based News Summarisation

Student Researcheracademic

Developed a transformer-based news summarisation system for processing and condensing changing multi-source news content; Capstone I: 91/100 (A), Capstone II: 90/100 (A). Worked on preprocessing, context management, redundancy across overlapping articles and quantitative model evaluation, strengthening end-to-end NLP experimentation skills. The project motivated later work on retrieval, evidence grounding and efficient handling of large or continuously changing information sources.

PythonNLPTransformersFine-tuning

InsightForge — Retrieval-Augmented Document Intelligence

Lead Developerpersonal

Built an end-to-end RAG application over PDF, DOCX and CSV collections using document parsing, chunking, embeddings, vector search and LLM-based grounded response generation. Designed ingestion and retrieval pipelines with pgvector similarity search, metadata-aware source handling and reranking to improve relevance across heterogeneous documents. Implemented streaming, citation-aware answers and source traceability so generated responses could be checked against supporting evidence instead of treated as opaque output. Exposed the workflow through FastAPI and containerised the system with Docker, creating a reproducible foundation for deployment, testing and further product iteration.

PythonFastAPILangChainGroq LLMPostgreSQLpgvectorDocker

Education

Bachelor of Science in Data Science

PUCIT, University of the PunjabData Science2021 – 2025

Thesis. BRIEFLY — Transformer-Based News Summarisation

Coursework

Artificial IntelligenceMachine LearningBig Data AnalyticsMachine Learning OperationsDatabasesProbability & StatisticsLinear AlgebraData Structures & AlgorithmsAnalysis of Algorithms

Recognition

Awards

Runner-Up — Coding Spree 2024

COMSATS2024

6th Place — NASCON 2024

FAST Islamabad2024

Gold Medal, 3rd Place — ICPC Asia Topi Regional On-site Contest

ICPC

Silver Medal, 8th Place — ICPC Asia Topi Regional Preliminary

ICPC

Runner-Up — PUCIT Programming Competition

PUCIT

Hire MuhammadReach out about a role, a contract or a conversation.For recruiters

Muhammad's twin is AI, it can make mistakes.

More people on Kanvis

Browse more people Create your own Kanvis