faizan ul hassan

faizan ul hassan

AI Full Stack Developer specializing in Agentic Workflows and RAG Pipelines

Based in: Islamabad, PakistanMost recently: AI Full Stack Developer, Xperion (Agentic Dream)Experience: 3 years

Hire me

Experienced AI Full Stack Developer with over 3 years of expertise in building production-grade LLM applications and multi-agent systems. Proven track record in delivering end-to-end GenAI solutions for hiring automation, legal-tech, and enterprise compliance using modern stacks like FastAPI, React, and AWS Bedrock.

Experience

AI Full Stack Developer

Xperion (Agentic Dream)work2026 – Aug 2026

Summary

Owned React/TypeScript frontends, Supabase backends, Dokploy deployments, and AWS Bedrock LLM workflows. Led two junior engineers; client-facing technical lead from requirements to stakeholder presentations.

What I did

  • DreamIT: Built a company management platform used by 250+ employees, unifying client/project delivery, time tracking, financials, and resourcing in company-wide and per-project dashboards, with 7-role access control and BambooHR, NetSuite, Jira, GitHub, and KnowBe4 integrations.
  • AI-Automated Hiring Pipeline: Built an AWS Bedrock pipeline that screens candidates and auto-schedules interviews, cutting time-to-decision from 2-3 weeks to 4-5 days. Led its Foloup (video interview) and Gorilla (coding assessment) modules, monitored via self-hosted Arize Phoenix.
  • POD Journey: Contributed to an agile team-management platform integrated with DreamIT, supporting 30-day constitution and 90-day Kaizen cycles with maturity assessments, DORA metrics, automated notifications, and performance monitoring.

Results

  • Cut time-to-decision in hiring from 2-3 weeks to 4-5 days.
  • Unified company-wide delivery and financials for 250+ employees.
ReactTypeScriptSupabaseDokployAWS BedrockBambooHRNetSuiteJiraGitHubKnowBe4Arize Phoenix

AI Backend Engineer

QuestworkJul 2024 – Dec 2025

Summary

Built backend systems, RAG pipelines, agentic and multi-agent workflows for client and internal projects. Traced LLM workflows with LangSmith; ran OpenEvals evaluations on internal tools and select client projects.

What I did

  • AI Courtroom Hearing Portal: Cut multi-week manual hearing processes to zero effort by leading backend development (FastAPI/Node.js) for 3 counties including video ingestion, speaker-identified transcripts, notes, action items, and an in-meeting chatbot across an average of 10 daily hearings, along with AI-generated newsletters on configurable schedules.
  • Multi-Agent Workflow Assistant: Reduced HR's exam-registration work to a single review step with a LangGraph copilot handling queries, registration, challans, payment proofs, and automated emails and slips for 9 monthly exams across 3 cities (500+ registrations each).
  • Academic Compliance Assistant: Reduced faculty review workload for UAE accreditation audits with an AI assistant analyzing 45-50 documents per course against Bloom's Taxonomy matrices for PLO/CLO alignment.
  • Agentic RAG for Multi-Hop QA: Reduced incomplete and inaccurate answers across production RAG bots with an agentic RAG framework (LangChain, LangGraph, ChromaDB) using multi-hop, chain-of-thought reasoning.
  • Built ETL pipelines to scrape, clean, and deduplicate content, normalizing it into structured documents for indexing.
  • Built separate FastAPI services for YouTube ingestion, web ingestion, and automated newsletter generation and dispatching.
  • Separated resource-intensive ingestion and AI processing into independent microservices to prevent blocking platform operations.
  • Implemented asynchronous FastAPI endpoints, API rate limiting, and error handling to manage external service constraints.
  • Persisted processing statuses and checkpoints in PostgreSQL with retry mechanisms for interrupted jobs.
  • Built an HR exam registration copilot using Python, LangChain, and LangGraph to manage registrations, fee challans, payment verification, and communications.
  • Used Claude and Gemini to generate structured content for meeting overviews, legal summaries, bill extraction, and newsletters.

Results

  • Achieved sub-100ms search latency for an enterprise semantic search platform indexing technical content from three websites.
  • Reduced HR exam-registration work to a single review step for 500+ registrations per exam.
FastAPINode.jsLangGraphLangChainChromaDBLangSmithOpenEvalsElasticsearchAnthropic ClaudeGoogle GeminiPostgreSQLPythonPydanticJSON schemas

Python Backend Developer

Codeaza TechnologiesworkJul 2023 – Jun 2024

Summary

Developed Python backend services and APIs, managed databases, and automated CI/CD deployments for client projects. Communicated directly with clients on project updates and delivery.

What I did

  • Sentimantle: Built backend pipelines processing hundreds of thousands of AI-curated Google, Booking.com, and TripAdvisor reviews for a unified hospitality sentiment dashboard.
  • Home Schooling Platform: Engineered a school management platform for an Australian home-schooling provider across 6 states, consolidating curriculum, student administration, and AI-driven performance reporting.
  • Work Book: Cut manual rulebook reviews (7.5 hrs/week per person) by designing the backend of an AI compliance-monitoring platform for investment banks, automating version comparison and plain-English action items.

Results

  • Processed hundreds of thousands of AI-curated reviews.
  • Cut manual rulebook reviews by 7.5 hrs/week per person.
PythonCI/CDGoogle Reviews APIBooking.com APITripAdvisor API

Skills

Technical

RAG Systems
Agentic & Multi-Agent Systems
OpenAI
LangGraph
PythonPython
LangChain
Google Gemini
Git & GitHubGit & GitHub
OpenEvals
AWS Bedrock
Arize Phoenix
FastAPIFastAPI
ReactReact
LLM-as-Judge Evaluation
LLM Observability
JavaScriptJavaScript
TypeScriptTypeScript
SupabaseSupabase
ChromaDB
LangSmith
REST APIs
Anthropic Claude
AWS
Node.jsNode.js
ViteVite
Tailwind CSSTailwind CSS
Ollama
Deepgram
PostgreSQLPostgreSQL
MySQLMySQL
MongoDBMongoDB
Pinecone
Dokploy
CI/CD
DockerDocker
Claude Code
n8n
WebSockets
Microservices
Docling
Elasticsearch
FlaskFlask
RabbitMQ
Google Cloud PlatformGoogle Cloud Platform
DigitalOceanDigitalOcean
Pydantic
Asynchronous Programming
Tool Use Orchestration
JSON schemas
Multi-agent workflows
Agentic RAG
Structured Outputs

Projects

Academic Compliance Assistant

Developerprofessional
An AI assistant analyzing 45-50 documents per course against Bloom's Taxonomy matrices for PLO/CLO alignment for UAE accreditation audits.
AIBloom's Taxonomy

Sentimantle

Backend Developerprofessional
Backend pipelines processing hundreds of thousands of AI-curated Google, Booking.com, and TripAdvisor reviews for a unified hospitality sentiment dashboard.
PythonAI curation

DreamIT

Lead Developerprofessional
A company management platform used by 250+ employees, unifying client/project delivery, time tracking, financials, and resourcing in company-wide and per-project dashboards, with 7-role access control and BambooHR, NetSuite, Jira, GitHub, and KnowBe4 integrations.
ReactTypeScriptSupabaseBambooHRNetSuiteJiraGitHubKnowBe4

AI-Automated Hiring Pipeline

Lead Developerprofessional
An AWS Bedrock pipeline that screens candidates and auto-schedules interviews, cutting time-to-decision from 2-3 weeks to 4-5 days. Includes Foloup (video interview) and Gorilla (coding assessment) modules, monitored via self-hosted Arize Phoenix.
AWS BedrockArize Phoenix

Multi-Agent Workflow Assistant

Developerprofessional
A LangGraph copilot handling queries, registration, challans, payment proofs, and automated emails and slips for 9 monthly exams across 3 cities.
LangGraph

Home Schooling Platform

Backend Developerprofessional
A school management platform for an Australian home-schooling provider across 6 states, consolidating curriculum, student administration, and AI-driven performance reporting.
PythonAI reporting

Work Book

Backend Designerprofessional
An AI compliance-monitoring platform for investment banks, automating version comparison and plain-English action items for rulebook reviews.
PythonAI Compliance

Agentic RAG for Multi-Hop QA

Developerprofessional
An agentic RAG framework (LangChain, LangGraph, ChromaDB) using multi-hop, chain-of-thought reasoning to reduce incomplete and inaccurate answers.
LangChainLangGraphChromaDBLangSmith

POD Journey

Contributorprofessional
An agile team-management platform integrated with DreamIT, supporting 30-day constitution and 90-day Kaizen cycles with maturity assessments, DORA metrics, automated notifications, and performance monitoring.
AgileDORA metrics

TechAI

personal
An enterprise semantic search platform that indexed technical content from three websites using custom mappings, analyzers, and fuzzy matching for typo-tolerant search.
Elasticsearch

AI Courtroom Hearing Portal

Backend Leadprofessional
Backend development (FastAPI/Node.js) for 3 counties including video ingestion, speaker-identified transcripts, notes, action items, and an in-meeting chatbot across an average of 10 daily hearings, along with AI-generated newsletters.
FastAPINode.jsAI TranscriptsAnthropic ClaudeGoogle GeminiPostgreSQLPydanticJSON schemas

Voice-based AI Dental Receptionist

personal
A voice-based agent designed to handle front-desk responsibilities including patient inquiries, appointment management, and administrative workflows.
PydanticJSON schemas

Education

Bachelor of Science

FAST-NUCESComputer Science2019 – 2023
Hire faizanReach out about a role, a contract or a conversation.For recruiters

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

More people on Kanvis

Browse more peopleCreate your own Kanvis