FARAIH NAZAR

Junior AI/ML Engineer | RAG Specialist | Building AI for Legal & Business Automation

Based in: Lahore, PakistanMost recently: Intern, ISPR (Inter-Services Public Relations)Experience: 1 year

Hire me

Junior AI/ML Engineer specializing in RAG architectures and automated data pipelines. I've built production-grade systems ranging from multi-agent legal assistants for Pakistani law to live WhatsApp business chatbots. With a background in architecting resilient ETL pipelines and deploying ML models for demand forecasting, I bridge the gap between complex data and actionable AI solutions. Proven leader through roles at UETXNews and the New York Academy of Sciences.

Experience

Intern

ISPR (Inter-Services Public Relations)Pakistanwork

Summary

Assisted with media and communications projects.

What I did

  • Worked within Pakistan's primary military media and communications institution.

Junior AI/ML Engineer

Acceleration Limitedwork2025 – 2026

Summary

Architected and maintained RAG pipelines across multiple knowledge bases: document ingestion, chunking, embedding generation, vector indexing, and retrieval-grounded response generation. Built and evaluated ML models for pricing analysis and demand forecasting using feature-engineered datasets from travel and e-commerce sources. Engineered resilient web scrapers and ETL pipelines across 5+ platforms, handling pagination, layout drift, and schema variation to deliver consistent data feeds. Collaborated directly with stakeholders to translate business requirements into deployed APIs and measurable product improvements.

Results

  • Architected an automated normalization pipeline using OpenAI API to standardize schemas for 200,000+ product listings across 10+ brands, enabling consistent semantic search results.
RAGML modelsETLWeb ScrapingAPIsVector Indexing

Skills

Technical

Vector Search
WhatsApp Cloud API
Web Scraping
Prompt Engineering
PythonPython
Pandas
RAG
LangChain
Streamlit
Model Training & Evaluation
Feature Engineering
NumPy
Data Cleaning
GitGit
SentenceTransformers
SQLSQL
Embeddings
Retrieval Tuning
Agent Orchestration
Qdrant
FastAPIFastAPI
EDA
Pipeline Automation
ETL
Selenium
BeautifulSoup
REST APIs
scikit-learn
DockerDocker
MongoDB AtlasMongoDB Atlas
FAISS
Schema Standardization
Playwright
Matplotlib
Seaborn
Plotly
Crawlee

Projects

Conflict Detection & Resolution System

Developerpersonal
A web-based team collaboration tool that automatically detects and resolves project conflicts - covering resource contention, scheduling overlaps, dependency clashes, task duplication, and communication gaps - and surfaces AI-assisted resolution suggestions or allows manual resolution with full audit reports. Built an MCP integration layer providing unified SSO, MFA, and role-based access control across multiple project environments from a single central server.
MCP IntegrationSSOMFARBACAI-assisted resolution

Cognitive Learning Using AI

Researcher/Prototyperacademic
Drove prototyping and experimentation within a global student research team, turning AI-driven learning concepts into testable, working demos.
AI-driven learning

Verdict System: AI Legal Assistant for Pakistani Law

Developerpersonal
A web app where law students type a legal question in plain language and receive cited, context-grounded answers drawn from Pakistani legal texts. Engineered a custom OCR pipeline to digitize and index historical government legal documents, enabling RAG-based retrieval from sources that were previously inaccessible. Currently a research-focused prototype with a focus on expanding the indexed legal corpus.
Multi-agent RAGLocal LLMVector SearchStreamlitOCR

Fashion Discovery Engine (Semantic Search)

Developerpersonal
Built a web-based fashion search engine that lets users find clothing across multiple brands using natural language. Engineered an automated normalization pipeline using OpenAI API to standardize schemas for 200,000+ product listings across 10+ brands, ensuring consistent and accurate semantic retrieval.
Semantic SearchWeb ScrapingData Ingestion PipelineSimilarity SearchOpenAI APIETL

WhatsApp Business Knowledge Assistant

Lead Developerpersonal
A live WhatsApp chatbot on Meta's official Cloud API that provides automated, source-grounded answers to customer queries. Integrated with a fashion discovery engine to support semantic product searches (e.g., "show me black silk outfits"), citing sources in every reply. Supports full multi-turn conversation context and falls back to web search when internal knowledge is unavailable.
WhatsApp Cloud APIMeta Cloud APIRAGWeb Search IntegrationSemantic Search

Education

BS Computer Science

University of Engineering & Technology, LahoreComputer ScienceJun 2027

Societies & activities

UETXNews

Governance & advisory

Director

UETXNewsother

Led the interviews & podcast team, overseeing content direction and team coordination.

Team Lead

New York Academy of Sciences (NYAS)other2023 – 2024

Led a 5-member international student research team on AI-driven cognitive learning initiatives.

What drives the work

Motivations

  • Building reliable AI systems for legal and public sectors
  • Mastering RAG evaluation frameworks and AI reliability
Hire FARAIHReach out about a role, a contract or a conversation.For recruiters

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

More people on Kanvis

Browse more peopleCreate your own Kanvis