Open to opportunities & collaborations

Babar AliAI Engineer

>

I build reliable LLM agents, RAG systems and production AI — measured, guarded and deployed.

View projects Resume Multan, Pakistan
Babar Ali
● 6 AI systems live

6

AI systems deployed

1,000+

automated tests

12

verified certifications

2

industry AI roles

Live systems

Real deployments, checked right now. Free hosting sleeps when idle.

Selected work

Production AI systems, not notebooks

Six end-to-end systems, each deployed with tests, a frozen held-out evaluation and documented limitations. Open any project for the architecture and results.

Agentic AI

TicketPilot

Customer-support agent that can safely issue refunds

A tool-calling LLM agent that resolves e-commerce tickets end to end — refunds, address changes, order tracking — and drafts the reply, while deterministic code keeps control of the money.

7/7

injection attacks caught

0/74

false positives

$50

human-approval threshold

Python · LangGraph · FastAPI · PostgreSQL

LLM + Data

QueryWise

Governed text-to-SQL analytics assistant

Plain-English business questions become verified SQL, a table, a chart and a summary whose numbers are checked against the result — over a 45k-row e-commerce warehouse.

57/57

malicious queries blocked

0

false blocks

5 s

hard query timeout

Python · sqlglot · PostgreSQL · pgvector

Document AI

LedgerLens

LLM invoice extraction with human review

0

false auto-approvals

11/11

planted anomalies caught

Python · Gemini · FastAPI · PostgreSQL

RAG

PolicyPilot

Permission-aware RAG assistant

0

permission leaks

97.1%

citation precision

Python · pgvector (HNSW) · Gemini embeddings · FastAPI

AI Infrastructure

ToolGate

Security gateway for AI agents (MCP)

29

policy decision cases

33 ms

median overhead

Python · Model Context Protocol · FastAPI · PostgreSQL

IoT + ML

FreeSpot

Real-time IoT classroom availability (Final Year Project)

Shows students and staff, live on their phones, whether each room is free, a class is running, a teacher is late, or a class was cancelled or moved.

307 ms

p95 event-to-dashboard

234

backend tests

0

cameras or microphones

Python · ESP32 · mmWave radar · MQTT

For recruiters

Does Babar fit your role? Ask the AI.

Paste a job description. The AI compares it with Babar's real projects and experience and gives an evidence-based fit analysis — including honest gaps.

The analysis appears here: a fit score, the requirements Babar meets with evidence, and the gaps.

Experience

Where I've been building

AI Engineer · Sada Digital Technology

Jun 2026 — Present

Doha, Qatar · Remote

  • Build AI-powered automation and data-driven solutions integrated into custom client software, aligning technology with business objectives to enable smarter decisions and scalable growth.
  • Take AI features from business requirement to production: scoping the use case, designing the data flow and integrating ML models and LLM APIs into existing applications.

Data Scientist · Vakial Tech

Feb 2026 — Present

Multan, Pakistan

  • Build and deploy machine learning models for data analysis and business intelligence.
  • Develop interactive Power BI dashboards and AI automation pipelines that streamline recurring data workflows.

Campus Brand Ambassador · DataCrumbs

Jun 2025 — Sep 2025

MNS University of Agriculture, Multan

  • Promoted DataCrumbs' AI and Data Science live sessions and research projects across academic and tech communities.

B.S. Information Technology · MNS University of Agriculture, Multan

2023 — 2027 (expected)

Final Year Project: FreeSpot, an IoT classroom availability system.

How I build AI

Three rules every project follows

Measure first

Every project ships with a frozen held-out test set, run once and logged, so results can't be tuned to the test.

Model proposes, code decides

Deterministic checks own money, permissions and data access. The LLM never holds that authority.

Honest documentation

Each README reports limitations and failed experiments next to the results.

Toolkit

Skills

Generative & Agentic AI

LLM AgentsLangGraphLangChainMCPTool CallingRAGText-to-SQLStructured OutputsPrompt EngineeringHugging Face

Evaluation & AI Safety

Held-out EvaluationRecall@k / MRRAblation StudiesPrompt-Injection DefenseGuardrailsHuman-in-the-LoopAudit Logging

Machine & Deep Learning

Scikit-learnPyTorchTensorFlowKerasCNNsTransformersComputer VisionNLP

Data Science

PythonSQLPandasNumPyStatisticsPower BIData Visualization

Backend & Data

FastAPIPostgreSQLpgvectorSQLAlchemyPydanticRESTServer-Sent Events

Engineering & MLOps

DockerGitHub ActionspytestmypyOpenTelemetryMLflowRenderVercel

Credentials

Verified certifications

Each card links to the issuer's public verification page.

Contact

Let's build reliable AI together.

Hiring, looking to collaborate, or have an AI idea to bring to life? Reach out — I reply to every message.