Abdullah Saif

AI Product Engineer specializing in LLM integrations, RAG systems, and automated workflows.

Based in: Lahore, PakistanMost recently: Software Engineer, AI Products Development, DownlabsExperience: 3 years

Hire me

Abdullah Saif is an AI/ML Engineer with extensive experience building production-ready AI products, including security auditing platforms and automated knowledge base pipelines. He combines a strong foundation in Artificial Intelligence with practical expertise in Python, LangChain, and multi-model LLM architectures.

Experience

Software Engineer, AI Products Development

Downlabswork2024 – Aug 2026

Summary

Develop AI-driven products, including SiteDNA, an automated web security and penetration-testing platform. Integrate multi-AI systems using Gemini, Claude, and GPT-4o for code remediation, report generation, and content pipelines. Own product decisions end to end, including architecture, API design, deployment, and client-facing reporting.

What I did

  • Implemented parallel computing and multi-threaded processing to run different tasks on different threads.
  • Completed contract as a Software Engineer, AI Products Development.

Results

  • Reduced scan time from 18 seconds to approximately 3 seconds for web scans and 6 seconds total including report generation.
GeminiClaudeGPT-4oAIREST API

Associate Software Engineer

Artificial Automation Business SolutionsworkNov 2025 – Jul 2026

Summary

Build AI-powered features with Python and LLM APIs to automate business workflows. Develop and maintain production web application features using React and Python. Collaborate with an engineering team using Git, sprint planning, and peer code reviews.

What I did

  • Used AI to develop Android mobile applications for one year.
  • Developed a Google Maps scraper to collect leads and push them to the SiteDNA pipeline.
PythonLLM APIsReactGitAndroid Development

AI Intern

Superior Universityinternship2025 – Dec 2025

Summary

Built a chatbot from scratch using Python and NLP, trained on Islamic literature. Owned the pipeline from data collection and model training through deployment.

What I did

  • Controlled all development work and designed the architecture for the Medical Chatbot project.
PythonNLP

Skills

Technical

RAG
Prompt Engineering
PythonPython
LangChain
OpenAI
GitGit
REST APIs
FlaskFlask
Claude
PostgreSQLPostgreSQL
Gemini
Pinecone
Qdrant
NLP
AWS
JavaScriptJavaScript
React.jsReact.js
MySQLMySQL
Node.jsNode.js
MongoDBMongoDB
Parallel Computing
Multi-threading
GraphRAG
Neo4j
RBAC
Android Development
System Prompts
MCP
Schema Validation
FastAPIFastAPI
Ollama
Ablation
Quantization

Languages

English

Projects

SiteDNA

Lead Developerprofessional
Enterprise security auditing platform with 15+ parallel scan engines covering vulnerability, network/port, CMS/technology, SSL/TLS, and breach intelligence. Supports bulk scanning of 1,000+ domains, white-labeled PDF executive reports, and a REST API for CI/CD integration.
Next.jsGeminiClaudeGPT-4o

Medical Chatbot

Developerpersonal
Built an end-to-end RAG pipeline over a medical knowledge base with vector search, retrieval chains, and grounded answer generation. Deployed the Flask API on AWS.
RAGPythonFlaskLangChainPineconeAWSGraphRAGNeo4jPostgreSQLAgent harness techniques

Automated Knowledge Base Ingestion Pipeline

Developerpersonal
Built an ETL workflow that retrieves documents, splits text, generates embeddings, and upserts vectors into Qdrant for live RAG queries.
n8nGoogle Drive APIOpenAI EmbeddingsQdrant

AutoML Data Prep & Modeling

Developerpersonal
Built a no-code ML platform for preprocessing messy datasets, training multiple algorithms, and downloading trained models.
PythonStreamlitScikit-LearnPandasNumPy

LeadsIn

Developer/Founderpersonal
Built and sold a commercial lead-generation extension that researches leads with AI and drafts personalized outreach.
JavaScriptOpenAI APIChrome Extension APIsGoogle Maps

Code Guard

Developerpersonal
A software system that handles the process from code writing to production-ready status, including CI/CD checks similar to OpenShift.

MCP Servers

Developerpersonal
Currently developing Model Context Protocol (MCP) servers for agentic tool interaction.

Image Generation Pipeline

personal
An image generation pipeline where LLMs are enforced to follow specific inputs and outputs across two scenarios: reference image mode and creative mode.
System Prompts

Natural Language PostgreSQL Chatbot

Developerpersonal
Built a conversational agent that translates natural-language questions into SQL, executes queries against PostgreSQL, and returns answers.
n8nOpenAI GPT-4o-miniPostgreSQLJSON

Education

B.S. Artificial Intelligence

Superior University, LahoreArtificial IntelligenceGPA 3.32024

Recognition

Certifications

Microsoft Azure AI Fundamentals

Microsoft2025

Prompt Engineering for Developers

Vanderbilt University
Hire AbdullahReach out about a role, a contract or a conversation.For recruiters

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

More people on Kanvis

Browse more peopleCreate your own Kanvis