— bash — ali@portfolio:~
ali@dev:~$ whoami

Ali Umair

ali@dev:~$ cat role.txt
Generative AI Software Engineer · LLM Agents · RAG Systems
ali@dev:~$ locate me
Lahore, PK open to relocation · New York, NY
ali@dev:~$
portrait.jpg 1024 × 1084
Ali Umair
status: available for work
SCROLL

cat about.md

Software engineer with 6+ years building production SaaS, backend systems, and AI/LLM workflows.

I design multi-tenant systems, AI agent pipelines, RAG-style retrieval, embeddings workflows, LLM tool-calling, data processing pipelines, and human-in-the-loop AI product infrastructure.

Experienced across testing, debugging, documentation, deployment lifecycle management, and production reliability — Ruby on Rails, Node.js, TypeScript, Python, PostgreSQL, Redis, Docker, Kubernetes, AWS, and CI/CD environments.

6+
years shipping software
5
production AI / ML projects
4
companies shipped for

ls ~/skills --grouped

languages/

Python TypeScript JavaScript Ruby SQL Java / C++ (academic)

ai_ml_systems/

LLM agents RAG embeddings pgvector prompt engineering tool-calling model routing structured outputs

ml_infra_data/

data pipelines background jobs model-output eval structured extraction crawler signals analytics ingestion queue workflows

backend_cloud/

Ruby on Rails Node.js REST APIs GraphQL PostgreSQL Redis Sidekiq Docker Kubernetes AWS EC2

testing_quality/

RSpec integration testing API testing CI/CD GitHub Actions root-cause debugging security-aware dev

frontend/

React Next.js TypeScript Tailwind CSS Hotwire / Turbo

git log --experience --oneline

current

Software Engineer / AI Product Engineer

  • Build and maintain production AI-powered SaaS features for lead discovery, prospecting automation, engagement workflows, and data-driven decision support.
  • Develop backend services for multi-tenant data processing, automation scheduling, account-level isolation, and scalable background job execution.
  • Integrate LLM workflows into product features — prompt-driven analysis, structured outputs, automation rules, and human-review approval flows.
  • Design data pipelines for extracting, normalizing, scoring, and organizing prospect, account, and engagement signals used in AI-assisted workflows.
  • Support production reliability across Rails, PostgreSQL, Redis, Sidekiq, Docker, Kubernetes, AWS EC2, and GitHub Actions.

Senior Software Engineer

  • Built production-grade SaaS and AI-enabled application features across backend, frontend, and infrastructure layers.
  • Designed APIs, database schemas, background jobs, integration workflows, and implementation plans for client-facing software products.
  • Contributed to architecture, technical planning, debugging, documentation, and delivery using Rails, Node.js, React, PostgreSQL, Docker, and cloud environments.

Associate Software Engineer / Rails Developer

  • Developed backend and full-stack features using Ruby on Rails, PostgreSQL, REST APIs, JavaScript, and third-party integrations.
  • Implemented database-backed business logic, authentication flows, admin features, API endpoints, and customer-facing workflows.
  • Participated in code reviews, testing, deployment, debugging, and production support activities.

Web Developer

  • Built and maintained web applications, dashboards, landing pages, and client-facing product features for international clients.
  • Worked across frontend and backend using JavaScript, HTML/CSS, APIs, SQL databases, and server-side application logic.

ls ~/projects/ai-ml --featured

flagship

AI Receptionist SaaS

  • Production-style AI receptionist platform with LLM orchestration, RAG retrieval, embeddings, tenant-aware memory, and tool-calling workflows.
  • Hybrid retrieval combining vector + lexical search to improve answer grounding and cut hallucination risk.
  • Prompt builders, model routing, response validation, appointment-request flows, human handoff, and tenant-isolated knowledge ingestion.
  • Background ingestion pipelines for documents, chunking, embeddings generation, cache invalidation, and retrieval-ready storage.
Rails RubyLLM OpenRouter PostgreSQL pgvector Sidekiq ActionCable

RankPilot

AI SEO Agent Platform
  • AI SaaS workflow for SEO analysis using Search Console, Analytics, crawler data, deterministic planning, and human approval loops.
  • Ingestion and processing to analyze website signals, generate improvement plans, and support repeatable optimization tasks.
  • Focused on AI governance, structured outputs, and approval gates over uncontrolled autonomous execution.
Rails Search Console Analytics AI planning

PlacesAgent

ReAct-style tool-calling agent
  • AI agent workflow using LLM tool-calling, Google Places data, structured reasoning, and external API integration.
  • Separated reasoning, tool execution, and final response generation to improve reliability, traceability, and debuggability.
RubyLLM Google Places API ReAct

mempalace-rb

AI memory layer for LLM apps
  • Rails-native AI memory layer for LLM applications, retrieval workflows, and persistent AI context management.
  • RAG-compatible storage using PostgreSQL full-text search with optional pgvector support.
Ruby Rails full-text search pgvector

TradeHarness

Safety-first agent CLI harness
  • Local CLI harness where AI proposes actions and deterministic rules decide whether they are allowed.
  • Built around audit logs, local storage, policy enforcement, simulation-only execution, and manual instruction generation.
TypeScript CLI audit logs policy engine

cat education.txt

B.Sc. Computer Science

2018 — 2022
National College of Business Administration & Economics · Lahore, PK
Data Structures & Algorithms Database Systems Software Engineering Object-Oriented Programming Operating Systems Computer Networks Probability & Statistics Artificial Intelligence foundations
ali@dev:~$ ./contact.sh

Let's build something
intelligent.

Open to AI engineering, LLM infrastructure, and backend roles — including relocation to New York, NY.