MUHAMMAD AZAM

Senior Software Engineer | Agentic AI & Scalable Backend Architect

Based in: ISLAMABAD, PAKISTANCurrently: Senior Software Engineer, Aladdin B2BExperience: 12 years

Hire me

Senior Software Engineer specializing in Agentic AI and high-concurrency backend architectures. Proven track record of migrating legacy monoliths to scalable microservices and architecting autonomous agent workflows using LangGraph and CrewAI. Expert in RAG-based systems, real-time data processing, and complex fintech billing integrations for global platforms like Goonj and Capstartup.

Experience

Senior Software Engineer

Aladdin B2Bwork2022 – Present

Summary

Architected autonomous agent workflows using LangGraph and CrewAI, automating complex B2B matchmaking and multi-step decision-making processes. Optimized RAG-based systems for intelligent information retrieval, ensuring highly relevant, context-aware outputs for enterprise users. Deployed NLP-based classification models to automate sector tagging and company intelligence (BPA), improving data organization for millions of records. Built predictive models for attendee engagement, utilizing data-driven insights to optimize smart meeting scheduling and user experience. Led the development of a scalable backend using Nest.js, Express.js, Node.js, Python Flask, Laravel supporting real-time matchmaking and high-concurrency workflows. Managed third-party integrations (LinkedIn, Apollo, Zoom, Salesforce) and enforced architecture best practices while mentoring junior developers.

What I did

  • Replaced monolithic LLM agent prompts with a stateful agent graph and vector retrieval.

Results

  • Architected and deployed multi-agent orchestration using LangGraph and CrewAI for automated B2B matchmaking.
  • Designed and deployed a multi-agent NLP classification pipeline using LangChain to evaluate lead profiles and match B2B buyers in real time.
  • Reduced B2B lead matchmaking and intent-qualification processing time by ~65%.
  • Scaled throughput from hundreds to 10,000+ automated decision loops per day.
  • Achieved 92%+ retrieval/accuracy precision through asynchronous state graph execution and tool-calling optimization.
  • Increased instant accuracy to 92%+
  • Sped up processing time by up to 3X
LangGraphCrewAIRAGNLPNest.jsExpress.jsNode.jsPython FlaskLaravelLinkedInApolloZoomSalesforceLangChainpgvector

Software Engineer

MobinspireIslamabadwork2019 – 2022

Summary

Revamped a global ride-hailing platform (Capstartup), migrating legacy PHP monoliths to a scalable Node.js microservices architecture on AWS, resulting in 95% faster load times and 90% improved scalability. Engineered a large-scale cargo automation platform (FretBay) across Europe using the MEAN Stack, optimizing zone-based matching and live tracking via Socket.io to achieve 90% faster request matching. Designed a comprehensive healthcare scheduling system (GETShifa) using Laravel and MongoDB, enhancing operational efficiency by 80% with real-time calendar sync and Redis caching.

What I did

  • Optimized zone-based matching and live tracking via Socket.io.
  • Implemented real-time calendar sync and Redis caching.

Results

  • Achieved 95% faster load times and 90% improved scalability for Capstartup.
  • Achieved 90% faster request matching for FretBay.
  • Enhanced operational efficiency by 80% for GETShifa.
PHPNode.jsMicroservicesAWSMEAN StackSocket.ioLaravelMongoDBRedis

Software Developer

Digital Media Distribution MaxIslamabadwork2017 – 2019

Summary

Designed and implemented a scalable microservices architecture for a high-volume streaming platform Goonj, improving system uptime by 98% using Node.js, Express.js, and Python. Developed a robust CMS and multi-bitrate transcoding engine via FFmpeg, reducing content publishing time by 40% and ensuring seamless video delivery across web and mobile. Integrated Telenor and Easypaisa billing APIs, streamlining subscription workflows and increasing successful payment transactions for national-scale user bases. Optimized system responsiveness by 90% through Redis caching, Nginx load balancing, and advanced MongoDB indexing to handle concurrent data streams. Architected and deployed responsive frontends using React.js and Material UI, ensuring cross-platform compatibility and secure JWT-based API interactions. Streamlined infrastructure automation using Docker and Kubernetes on AWS and Google Cloud, reducing deployment time and enhancing system scalability by 99%. Built real-time communication layers and notification services using Socket.io and RabbitMQ to support live streaming and instant user engagement.

What I did

  • Optimized system responsiveness through Redis caching, Nginx load balancing, and advanced MongoDB indexing.

Results

  • Implemented Kafka for high-volume telemetry and event queuing from active streams.
  • Leveraged Apache Spark for batch and real-time aggregations for downstream analytics pipelines.
  • Improved system uptime by 98%.
  • Reduced content publishing time by 40%.
  • Enhanced system scalability by 99%.
  • Enabled real-time user engagement analytics at scale using Kafka and Spark.
Node.jsExpress.jsPythonFFmpegRedisNginxMongoDBReact.jsMaterial UIJWTDockerKubernetesAWSGoogle CloudSocket.ioRabbitMQKafkaApache Spark

Skills

Technical

RESTful APIs
Node.jsNode.js
PythonPython
AWS
LangGraph
Vue.jsVue.js
DockerDocker
CrewAI
React.jsReact.js
Multi-Agent Orchestration
MySQLMySQL
PostgreSQLPostgreSQL
MongoDBMongoDB
RAG (Retrieval-Augmented Generation)
RedisRedis
Express.jsExpress.js
LaravelLaravel
Microservices Architecture
Nest.js
Socket.io
TypeScriptTypeScript
FlaskFlask
PHPPHP
RabbitMQ
LangChain
Prompt Engineering
Vector Databases (Pinecone, Qdrant)
NLP
Predictive Analytics
Kafka
Distributed Systems
ETL/ELT Pipelines
Google Cloud (GCP)Google Cloud (GCP)
KubernetesKubernetes
CI/CD Pipelines
MS SQL ServerMS SQL Server
Elasticsearch
Material UIMaterial UI
BootstrapBootstrap
AngularJSAngularJS
Apache Spark
Recommendation Engines
Geospatial Clustering
FastAPIFastAPI
Oracle
Infrastructure as Code (IaC)
Azure
SvelteJSSvelteJS
Computer Vision
QLoRA
Llama
Mistral
Fine-tuning
pgvector
LoRA

Areas of expertise

Teamwork
Ownership
Innovation
Problem Solving
Leadership
Consulting

Projects

Capstartup

Software Engineerprofessional
Global ride-hailing platform revamped by migrating legacy PHP monoliths to a scalable Node.js microservices architecture on AWS.
PHPNode.jsMicroservicesAWS

FretBay

Software Engineerprofessional
Large-scale cargo automation platform across Europe using the MEAN Stack, optimizing zone-based matching and live tracking via Socket.io.
MEAN StackSocket.io

GETShifa

Software Engineerprofessional
Comprehensive healthcare scheduling system using Laravel and MongoDB, featuring real-time calendar sync and Redis caching.
LaravelMongoDBRedis

Goonj

Software Developerprofessional
High-volume streaming platform with a scalable microservices architecture and a robust CMS with multi-bitrate transcoding engine via FFmpeg.
Node.jsExpress.jsPythonFFmpegRedisNginxMongoDB

WinAds

Software Developerprofessional
Multi-platform pixel management system for Facebook, TikTok, and Snapchat, featuring sophisticated behavioral tracking and audience segmentation.
Facebook PixelTikTok AdsSnapchat Ads

Urgency Sales Booster

Software Developerprofessional
Conversion optimization toolkit featuring real-time location-based messaging and sticky checkout components for Shopify merchants.
Shopify

Cyrano

Software Developerprofessional
Social media management platform featuring robust role-based access control (RBAC) and automated usage-based billing via Stripe/Braintree.
LaravelVue.jsStripeBraintree

Education

Master of Computer Science (MCS)

Pir Mehr Ali Shah Arid Agriculture UniversityComputer Software EngineeringAug 2015

Bachelor of Science (BSc)

The Islamia University of BahawalpurComputer ScienceJun 2013

Recognition

Awards

Built and deployed ML models processing millions of records daily

Professional Achievement

Predictive insights generation

Optimized data pipelines and ML datasets

Professional Achievement

Cutting training time by 40%

Automated data validation and drift detection

Professional Achievement

Boosting data reliability by 30%

Hire MUHAMMADReach out about a role, a contract or a conversation.For recruiters

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

More people on Kanvis

Browse more peopleCreate your own Kanvis