Maede Jalali

Hey, I'm Maede Jalali

AI Engineer

Building agentic systems, multimodal tooling, and production AI products.

About Me

Applied AI Engineer

Applied AI Engineer with 5+ years of experience, driving AI development across backend, agentic workflows, and customer-facing solutions. I’ve led the end-to-end deployment of a conversational Travel Assistant integrating OpenAI, LangChain, LangGraph, and RAG pipelines. With a strong Python foundation, I’ve productionized multimodal and agentic AI systems, collaborated with stakeholders ranging from product to research, and worked directly with users to fine-tune outcomes. I bring a customer-centric mindset, deep technical intuition, and a track record of delivering impactful GenAI applications that bridge cutting-edge research with real-world value.

Location
Tehran, Iran
Phone
+98 (933) 426-5571
Email
majj2000.mail@gmail.com
Career Path

My Professional Journey

AI Engineer

Sindibad

Jan 2024 – Present

Dubai, UAE (Remote)

Acted as the thread lead for Sindibad’s AI Assistant initiative, coordinating across product, design, engineering, and support teams. Led roadmap planning, prioritized AI feature development, and aligned technical execution with user and business needs. Presented the Assistant in multiple investor demos and collaborated on support automation tools that doubled triage speed.

  • LLM Agent Systems & Tooling: Developed a production-grade AI assistant using OpenAI Agents SDK and LangGraph. Built a modular agent framework with over 10 tools including travel search, booking planner, cost estimators, and custom web search. Introduced a Coordinator Agent pattern for tool orchestration and fallback handling, boosting tool success rates by 20% and enabling scalable reasoning workflows.
  • Retrieval-Augmented Generation (RAG): Built a domain-specific RAG pipeline grounded in Sindibad’s business knowledge base, user orders, and real-time system state. This enabled context-aware responses across travel planning and support tasks. Improved knowledge recall accuracy from ~60% to 90%, reducing support ticket rates by 20% and significantly improving first-contact resolution.
  • Prompt Engineering & Evaluation: Iterated on prompts using human-in-the-loop testing, comparing GPT and Gemini models. Built a DeepEval-based pipeline to benchmark BLEU, ROUGE, and BERTScore—enabling automated testing with Slack alerts on prompt changes.
  • User Feedback & Conversation UX: Analyzed anonymized user chats, session replays, and direct feedback to optimize assistant tone, persona, and error recovery flows. These improvements increased successful order completion and reduced bounce rates across funnels.
  • Multimodal Input & Vision Features: Added support for voice recording and image input (passport OCR via Azure Vision + OpenAI), reducing booking friction and cutting rectification time by 35%.
  • Monitoring & Reliability: Integrated OpenAI Traces, Sentry, and Kibana for observability across agent flows and system errors.

Software Engineer

Alibaba

Apr 2022 - Jan 2024

Tehran, Iran (On-site)

Built and maintained high-scale mobile travel features with a focus on product impact and reliability.

Software Developer

Miare

Dec 2020 – Apr 2022

Tehran, Iran (On-site)

Engineered scalable features for Miare’s courier app, improving performance and support tools.

Skills & Expertise

Technical Proficiency

Python
OpenAI Agents
OpenAI APIs
OpenAI Evaluation
Gval
DeepEval
Agentic Architecture
LangGraph
LangChain
Azure Document Intelligence
Django
Async Programming
API Design
GitHub Actions
PostgreSQL
Redis
REST APIs
Featured Projects

Notable Work

AI Travel Telegram Bot

Delivered the end-to-end creation of an AI-powered Telegram bot, enabling users to seamlessly book flights and hotels via LLMs.

Key Features

  • Reduced end-to-end booking time from up to 10 minutes to under 1 minute compared to the main app
  • Integrated OpenAI-powered agents to handle user queries, plan trips, and finalize bookings
  • Enhanced user experience through natural language processing and conversational interface
  • Real-time integration with flight and hotel booking systems

Technologies Used

OpenAIPythonDjangoAI AgentsLangChainLangGraphTelegram APIRAG Systems

IoT Airport Lighting Control System

Delivered a mobile-based control system for airport apron lighting, enabling operators to adjust settings in just 10 seconds.

Key Features

  • Real-time control and monitoring of airport lighting systems
  • Intuitive mobile interface for quick adjustments
  • Secure IoT communication protocols
  • Efficient operator workflow optimization

Technologies Used

AndroidKotlinIoT ProtocolsReal-time SystemsMobile UI/UX

Chat with My AI Assistant

Feel free to ask my AI assistant any questions about my experience, skills, or background.

Get in Touch

Contact Information