Skip to main content
00IndexAvailableFSD 00:00 PKT

AbdullahGrewal.

N°01CurrentlySenior AI/ML Engineer @ Infinitiv AIN°02LocationLahore, Pakistan / UTC+5N°03Focus▸▍N°04StatusOpen to new work
See my work →Get in touch
Plate · 01
Edition26
End of introScroll
01SectionProfileBio · Index

AboutMe.

N°01ProfileSeven tiles · Bio, Now, Stats, Location, Résumé, Mark
Abdullah
Available for Work

About

AI/ML Engineer who ships products.

Passionate about AI research, Deep Learning, and NLP. I build and fine-tune transformer models, optimize deep learning architectures, and ship full-stack AI applications.

Currently

Senior AI/ML Engineer

@ Infinitiv AI

0

Projects

0

Yrs AI/ML

0

Awards

Location

Lahore, Pakistan

UTC+5 · works async

Download résumé

PDF · 220 KB

Mark · 02
Signed’26
02SectionStackToolkit

SkillsSet.

N°02Stack4 categories · AI/ML, Web, Cloud, Tools
01/ 04

AI/ML & NLP

PythonMachine LearningDeep LearningNLPTransformersTensorFlowPyTorchLangChainScikit-learnFAISSChromaDBRAGHugging FaceOpenCVYOLOCrewAIWhisperDeepgram
02/ 04

Full Stack

ReactNext.jsNode.jsExpressFlaskFastAPIDjangoLaravelTypeScriptTailwindMongoDBPostgreSQLMySQLREST APIsGraphQLReact Native
03/ 04

DevOps & Cloud

DockerAWS S3AWS EC2Elastic IPAzureCI/CDGitHub ActionsLinuxUbuntuGitNginxVercel
04/ 04

Tools & Platforms

VS CodePostmanJupyterOllamallama.cppStripeFirebaseRedisPrismaDrizzleWebRTCFigmapgvector
InterludeProcessBuild · 4 Phases

How IBuild.

N°—BuildPrompt → Architecture → Code → Deploy
  1. 01/ 04

    Prompt

    User describes the problem.

  2. 02/ 04

    Architecture

    Map flows, pick models.

  3. 03/ 04

    Code

    Ship API + UI, wire RAG.

  4. 04/ 04

    Deploy

    Containerize, observe, iterate.

03SectionWorkTrack · Index

WorkLog.

N°03Track04 positions · 2023 → Now
01/ 042025 — Present

Senior AI/ML Engineer

@ Infinitiv AI

Lahore, PakistanHybrid

  • Designed end-to-end RAG-powered AI platforms and multi-agent systems.
  • Built production RESTful APIs with Flask and FastAPI.
  • Integrated ASR (Whisper, Deepgram) and TTS pipelines for real-time speech-to-speech systems.
  • Led frontend + Azure cloud deployment with Docker + CI/CD.
02/ 042024 — Present

Co-Founder & Lead AI Developer

@ MyndEdge

Lahore, PakistanHybrid

  • Built and scaled three investor-acquired AI products incl. AI-Powered Learning Platform and Maternity Care Chatbot.
  • Led end-to-end product lifecycle from ideation to deployment and investor handoff.
03/ 042025 — 2025

Associate AI/ML Engineer

@ Code District

Lahore, PakistanOnsite

  • Spearheaded backend with Python (Flask, FastAPI), improving system reliability by 45%.
  • Integrated AI features into Laravel projects, cutting manual workflows by 40%.
04/ 042023 — 2025

Senior Full Stack AI Engineer

@ NexoCore

Islamabad, PakistanHybrid

  • Built AI-powered web apps with Flask, React, Node, TensorFlow, PyTorch, LangChain.
  • Deployed scalable solutions on Azure and Heroku with Docker, CI/CD, Redis, RAG.
04SectionWritingFeed · Index

WritingNotes.

N°04Notes10 articles · Published on Medium
01/ 10Featured ▸

Apr 28, 202611 min read

Beyond Next-Token: How Diffusion LLMs Like Mercury 2 and LLaDA Hit 1,000+ Tokens Per Second in 2026

Read on Medium

§Excerpt · Opening
For five years, every flagship LLM has generated text the same way: one token at a time, left to right. In 2026, a new generation of diffusion language models is breaking that contract, and pushing…
Long-formFiled under Writing
02/ 10

Feb 17, 202613 min read

I Gave My Code to a Swarm of AI Agents and Here’s What Happened

Read on Medium

03/ 10

Nov 22, 202511 min read

Antigravity: Google’s New Agent-First IDE — How It Compares to Cursor and Kiro (Deep-dive…

Read on Medium

InterludeActivityGitHub · Index

Days ICode.

N°—ActivityLoading · @buzzgrewal
Year
Total—Days active—Longest streak—Best day—
Plate · HeatLoading
JanFebMarAprMayJunJulAugSepOctNovDec
MonWedFri
View on GitHub
LessMore
05SectionProjectsField · Index

ProjectsField.

N°05Field00 repositories · Live from GitHub · @buzzgrewalLive
Filter
06SectionAI LabLive · Demos

AILab.

N°06Lab05 demos · Streaming · Powered by OpenRouter + pgvectorLive
01/ Project Match

Tell me what you need.

I'll find Abdullah's projects that fit.

Try
N°—Flow
  1. 01Describe the problem
  2. 02Match by intent
  3. 03Ranked results
ResultsRanked by fit
04/ ChatbotOnline

BuzzBot

Persona-tuned. Ask about projects, stack, hiring.

Best project?Hiring me?RAG stack?Tell me about Whisper

How do I hire you?

You · 0:01

Open to roles + freelance. Tell me the stack, timeline, and scope — I'll point you to the right work.

BuzzBot · 0:02
Open chatBottom-right
02/ Code WhisperIdle

Paste code, get the gist.

Functions, snippets, configs, regex — streaming plain-English breakdown.

Try
N°—Languages
01TypeScript02JavaScript03Python04SQL05Bash06Regex07YAML08JSON09Go10Rust11GraphQL12Java13C++14C#15Ruby16PHP17Swift18Kotlin19HTML/CSS20Markdown21Dockerfile
N°—What it does
  • 01Plain-English summary
  • 02Spot likely bugs
  • 03Edge-case hints
  • 04Refactor pointers
StreamingLatencyToken-by-token
05/ NeuralOnline

AI Stack

Inference, embeddings, vector search — wired end-to-end.

N°—Models
01LLM Models02Embedding Models03Vector Database04Streaming Layer05Multi-Agent Framework06Speech Models
Streaming~312ms p50
03/ RAG Playgroundpgvector

Ask the corpus directly.

Embeds your question, searches pgvector, returns sources + answer.

Try
07SectionContactTouch · Index

Get inTouch.

N°07TouchDirect line · Email · Phone · FormAvailable
Plate · Direct
N°01

Email

abdullahramzan120@gmail.com

N°02

Phone

+92 306 9874952

N°03

Location

Lahore, Pakistan

N°—Elsewhere
01GitHub02LinkedIn03Medium04X / Twitter
N°—Local timeUTC+5 · PKT
––:––––
Off-hours

Mon — Sat · 09:00 → 22:00 PKT

Response~24h
Plate · Form

Send a message.

Project inquiry, hiring, collaboration — fastest channel for written context.

Encrypted · No tracking