Build record
AI Portfolio
An interactive AI portfolio website that uses LangChain and Pinecone to provide real-time RAG answers about Ameer's professional experience and system architecture.
- Next.js
- TypeScript
- RAG
- Langchain
- OpenAI
- Pinecone
Systems architecture · Reliable by design
I design systems that keep complex products moving.
I help teams turn difficult product requirements into scalable backend platforms, observable AI workflows, and cloud systems that remain dependable as they grow.
See the systemsSystem atlas / live model
Active routeA production system, end to end
ORCHESTRATION / SELECTED
Event-driven workflows coordinating long-running operations.
Proof by numbersVerified from professional history
7+years building production systems
99.2%crash-free sessions achieved
500K+learners served by Edkasa
Selected build records
Systems shaped around real constraints.
Each project joins architecture, delivery, and operational thinking—because the difficult part is rarely writing one service. It is making the whole system hold together.
Build record
DocuChat
Document-based chat application for enterprise teams featuring Role-Based Access Control (RBAC), multi-tenant isolation, and vector search over internal knowledgebases.
- Node.js
- Express.js
- React.js
- AWS
- TypeScript
- RAG
Build record
Fin Co-Pilot
Financial intelligence co-pilot providing fine-grained RBAC with OPA security layers for enterprises to query and analyze private datasets.
- Python
- Flask
- Node.js
- RAG
- LangGraph
- PostgreSQL
Build record
Edkasa Streaming
High-scale online learning and video streaming platform with HLS transcoding and real-time student engagement features.
- Node.js
- PostgreSQL
- AWS S3
- CloudFront
- FFmpeg
- Android
Build record
TAGMU
Automated ML computer vision system for livestock identification, speeding up insurance underwriting and claim verification.
- Python
- Node.js
- Machine Learning
- Image Recognition
- MongoDB
System inventory
A stack is useful when the choices make sense together.
I work across architecture, backend, AI, and infrastructure so important decisions do not disappear between teams.
Architecture
- System Architecture
- Microservices
- Event-driven systems
- REST APIs
- GraphQL
AI systems
- RAG
- LangGraph
- LangChain
- Text-to-SQL
- Vector Search
- Pinecone
Platform
- Node.js
- NestJS
- Python
- PostgreSQL
- MongoDB
- RabbitMQ
Delivery
- AWS
- Docker
- CI/CD
- Observability
- CloudFront
- ECS Fargate
Operating history
Seven years of widening the system boundary.
From mobile reliability to backend platforms and AI orchestration, each role expanded the scale of the problems—and the responsibility for the outcome.
01FAMS by Falkenherz
Software Architect & Full-Stack Engineer
FAMS by Falkenherz
Software Architect & Full-Stack Engineer
- Lead system architecture for microservices telematics & IWMP TADWEER enterprise waste management platforms.
- Architected smart recurring plans system with automated crons, driver scheduling lifecycle, and real-time event-driven compliance engine.
- Built production-grade Text-to-SQL RAG system enabling real-time database intelligence via natural language queries.
Production-grade natural language to database intelligence pipeline
TADWEER waste management with real-time compliance engine
Automated crons, driver lifecycle & event-driven orchestration
- NestJS
- Express.js
- PostgreSQL
- LangGraph
- Text-to-SQL
- RabbitMQ
- Microservices
02Opsin
Backend Engineer & Solutions Architect (AI/ML)
Opsin
Backend Engineer & Solutions Architect (AI/ML)
- Architected AI-powered RAG pipeline for automated financial data analysis with fine-grained RBAC.
- Designed and implemented multi-agent orchestration pipelines using LangGraph for multi-step reasoning and tool execution.
- Scaled microservices on AWS ECS Fargate with RabbitMQ messaging, OCR semantic search, and Open Policy Agent security.
Multi-step reasoning and tool execution orchestration
AI-powered pipeline with fine-grained access control
Fargate microservices with RabbitMQ & OPA security
- Python
- Flask
- Node.js
- LangGraph
- LangChain
- RAG
- AWS ECS
- RabbitMQ
03OneByte
Software Engineer
OneByte
Software Engineer
- Integrated LangChain & LlamaIndex RAG pipelines with Pinecone and pg-vector for high-precision vector retrieval.
- Developed standalone ML API for livestock image identification to automate insurance claim processing.
- Built and scaled EdTech REST APIs with Node.js/Postgres, JWT authentication, and automated CI/CD pipelines.
LangChain + LlamaIndex with Pinecone & pg-vector
Livestock identification for insurance automation
Node.js/Postgres APIs with JWT & automated pipelines
- Node.js
- PostgreSQL
- Pinecone
- pg-vector
- LangChain
- LlamaIndex
- Docker
04Edkasa
Software Engineer (Mobile & Backend)
Edkasa
Software Engineer (Mobile & Backend)
- Engineered video lecture streaming pipeline using AWS S3, CloudFront CDN, and FFmpeg server-side HLS transcoding.
- Built core Node.js/Postgres APIs and cross-platform mobile apps for high-traffic student base.
AWS S3 + CloudFront CDN with FFmpeg transcoding
Node.js/Postgres APIs serving high-traffic student base
- Node.js
- PostgreSQL
- AWS S3
- CloudFront
- FFmpeg
- Android
- iOS
05LOGICON
Android Developer
LOGICON
Android Developer
- Managed team of 4; improved crash-free user sessions from 92% to 99.2% via deep crashlytics profiling.
- Reduced application binary footprint by 25% through WebP image optimization and modular clean architecture.
Up from 92% via deep crashlytics profiling
WebP optimization & modular clean architecture
- Android
- Kotlin
- Java
- Firebase
- Clean Architecture
Next system / Open channel
Bring me the difficult part.
If you are shaping a backend platform, untangling an AI workflow, or preparing a system for its next stage of growth, I would like to understand the problem.