Abdul Rahim
Dubai, United Arab Emirates | +971 54 431 7789 | [email protected] | aryem.dev | linkedin.com/in/aryem
Professional Summary
Senior Software Engineer shipping production agentic systems — voice agents, adaptive tutors, hybrid graph + vector retrieval, and CI-gated LLM evaluation pipelines — to enterprise customers. 8+ years of full-stack engineering (TypeScript, React, Node.js) underpinning hands-on depth in LLM orchestration, RAG, agent memory, red-team hardening, and AI observability. Track record of taking AI features from research prototype to revenue-impacting product.
Experience
Senior Software Engineer | Easygenerator | Dubai, UAE Jan 2022 – Present
Lead engineer on the EG-Labs AI initiative, driving R&D from prototype to production while continuing to ship core-product features used by enterprise customers.
AI-Native Knowledge Platform (current)
- Architected an AI-native knowledge platform with 700+ third-party integrations (Nango + custom connectors), unifying company knowledge into a hybrid Neo4j (entity graph) + Pinecone (semantic vectors) retrieval layer that enables multi-hop reasoning across organizational data.
- Built meeting bots and a conversational interview agent (Recall + custom orchestration) that capture tacit knowledge from calls and structure it back into the graph as entities and relationships.
- Designed an agentic AI tutor with adaptive learning paths backed by long-term agent memory (Mem0) and a generative UI layer that streams JSON-rendered React components, composing lessons per learner.
EasyCoach — AI Roleplay & Coaching Platform (beta, several hundred customers)
- Stood up a production LLM evaluation pipeline with Langfuse (LLM-as-judge, regression suites, golden datasets) that caught a 20% role-reversal failure rate and surfaced fourth-wall breaks, prompt injection, off-topic drift, and multi-turn degradation before they reached users.
- Integrated Promptfoo into CI as a deploy gate — every prompt and model change runs against jailbreak red-team suites and multi-persona benchmarks before merge, eliminating safety and persona regressions from production.
- Shipped an AI recording-analysis and skill-library feature that scores audio/video coaching sessions on delivery, word choice, content structure, and skill proficiency against both static and dynamically generated competency rubrics.
Earlier EG-Labs & Core Product Work
- Prototyped real-time AI voice agents using LiveKit, Pipecat, and OpenAI, plus an automated multimodal content-generation pipeline on Vertex AI.
- Engineered a security overhaul for media assets using signed URLs and access authorization — a decisive factor in passing penetration testing and closing a major enterprise contract.
- Led development of an in-app media suite with video/audio recording and editing, increasing user-generated content by 30%.
- Shipped an AI voiceover feature on AWS Language AI, lifting feature adoption by 20%.
- Introduced the Architecture Decision Record (ADR) pattern across the squad and mentored engineers via code review, contributing to a 15% reduction in bugs.
Software Engineer | Unlayer | Lahore, Pakistan Aug 2021 – Jan 2022
- Built the “Workspaces” feature for multi-project management; migrated billing from Stripe.js to Chargebee.js on TypeScript, Node.js, and AWS Lambda.
Software Engineer | Constellation Software | Lahore, Pakistan Nov 2019 – Jul 2021
- Led frontend for Target CRM (TypeScript, React, Redux, Material-UI) and built a real-time SMS chat micro-service (React, Azure Functions, Twilio, SignalR) plus an SMS survey service on .NET Core.
Software Engineer | TetraHex | Lahore, Pakistan Jul 2017 – Oct 2019
- Improved client site performance by 80% via SSR, lazy loading, and Webpack tuning (Vue/Nuxt); built an image-resizing service that cut server storage 50% and a serverless app on AWS Lambda + DynamoDB + Cognito.
Skills
- AI / LLM Engineering: LLM orchestration, agentic systems, RAG, hybrid graph + vector retrieval, agent memory (Mem0), generative UI, multi-turn evaluation, red-teaming, prompt hardening
- AI Tooling: OpenAI, Vertex AI (multimodal), AWS Language AI, ElevenLabs, LiveKit, Pipecat, Cartesia, Langfuse, Promptfoo, Nango, Recall
- Vector & Graph: Pinecone, Neo4j, embeddings, semantic search, entity-relationship modeling
- Frontend: TypeScript, JavaScript (ES6+), React, Redux, Vue.js, HTML, CSS/LESS, Tailwind CSS, Material-UI, Webpack
- Backend: Node.js, NestJS, Express.js, .NET Core, Python (Django/Flask), Ruby on Rails, REST, GraphQL
- Languages: TypeScript, JavaScript, C#, Python, Ruby, Go
- Cloud & DevOps: AWS (Lambda, S3, DynamoDB), GCP (Vertex AI, Firebase), Azure (Functions, App Service), Docker, Kubernetes, GitHub Actions
- Databases: PostgreSQL, MongoDB, MySQL, Snowflake, Neo4j
Education
B.Sc. in Computer Science National University of Computer & Emerging Sciences | Lahore, PK | 2017