AI product & innovation · Tokyo

AI that works.
Impact that lasts.

I’m Ahmed, a telecom and AI leader and inventor. I turn complex problems into useful AI products through hands-on innovation.

Ahmed Bahaaeldin Darwish
16+
years building & leading
3
USPTO patents
9,600
hours saved annually

Ask about Ahmed

Grounded in his portfolio

What would you like to explore?

AI answers from portfolio content. Chat saved on this device and server.

Ask about Ahmed

Grounded in his portfolio

What would you like to explore?

AI answers from portfolio content. Chat saved on this device and server.

Selected work

Built to make a difference.

A selection of AI products, systems, and tools I’ve brought to life.

Telora — AI Research for Telecom

Telecom / AI Agents

Telora — AI Research for Telecom

Read project details

Designed and built Telora, a personal AI agent for telecom research. It brings telecom news, visual learning and continuous discovery into one workspace, supporting the exploration of fresh telecom ideas and the preparation of articles and patent drafts.

  • Telecom
  • AI Agents
  • Research
  • Visual Learning
  • Innovation
Visit Telora ↗
Pulse — Open Source

Open Source / MCP

Pulse — Open Source

Read project details

Built Pulse, an open-source hub for AI agents and the people working with them. MCP tools connect knowledge, projects, notes, todos, reminders, memory and agents in one place, across devices and platforms.

  • Open Source
  • MCP
  • AI Agents
  • Knowledge
  • Project Management
  • Memory
View project on GitHub ↗
BM Agent — Benchmarking Agent

Rakuten Mobile / AI Agents

BM Agent — Benchmarking Agent

Read project details

Built BM Agent, an AI agent harness for analyzing radio log files at Rakuten Mobile.

  • Rakuten Mobile
  • AI Agents
  • Radio Log Analysis
IdeaSketch SaaS Product

SaaS / LangGraph

IdeaSketch SaaS Product

Read project details

Built IdeaSketch, a production-ready SaaS application powered by LangGraph for intelligent workflow orchestration. Deployed using LangGraph Studio with a modern Vite.js frontend, leveraging Supabase for database management and authentication. Key skills: LangGraph architecture, SaaS development, Vite.js, Supabase, authentication systems, database design, and full-stack development.

  • SaaS
  • LangGraph
  • Vite.js
  • Supabase
  • Authentication
  • Full-Stack
  • AI Workflow
Advanced AI Chatbots using No-Code Tools

Flowise / n8n

Advanced AI Chatbots using No-Code Tools

Read project details

Designed and deployed advanced AI chatbots leveraging no-code platforms including Flowise, n8n, and Langflow. Demonstrated expertise in workflow automation, prompt engineering, and rapid prototyping for business solutions. Key skills: No-code AI development, workflow design, system integration, conversational AI, and intelligent automation.

  • Flowise
  • n8n
  • Langflow
  • No-Code
  • Prompt Engineering
  • Automation
  • Conversational AI
Advanced AI Chatbots with Local LLMs

System Design / Architecture

Advanced AI Chatbots with Local LLMs

Read project details

Designed and built AI chatbots running on small, locally hosted LLMs for privacy-sensitive and cost-constrained deployments. Tight prompt engineering and resource-efficient architecture let constrained models do real work — no cloud dependency, no data leaving the perimeter. Key skills: System design, local LLM deployment, prompt engineering, edge AI, resource optimization, privacy-focused architecture.

  • System Design
  • Architecture
  • Local LLM
  • Prompt Engineering
  • AI Integration
  • Edge AI
Automatic DriveTest Route Creation

FastAPI / Python

Automatic DriveTest Route Creation

Read project details

Built an automated system for DriveTest route generation using FastAPI, Python, and OpenMaps. Streamlined geospatial data processing and backend API development for telecom RAN optimization. Key skills: FastAPI development, Python automation, OpenMaps integration, DriveTest methodology, mobile network optimization, geospatial analysis, and API design.

  • FastAPI
  • Python
  • OpenMaps
  • DriveTest
  • RAN
  • Mobile Network
  • Geospatial
  • API
YOLOv10 Site Manager Image Detection

YOLOv10 / Computer Vision

YOLOv10 Site Manager Image Detection

Read project details

Developed an end-to-end computer vision pipeline for site management and safety compliance monitoring. Implemented real-time object detection and safety violation identification for telecom site operations. Key skills: YOLOv10 model deployment, computer vision, FastAPI development, image processing, and safety compliance automation.

  • YOLOv10
  • Computer Vision
  • FastAPI
  • Image Processing
  • Safety
Health and Safety AI Detection

YOLOv9 / Computer Vision

Health and Safety AI Detection

Read project details

Developed an AI-powered computer vision system for workplace health and safety compliance monitoring. Implemented automated detection of safety violations and compliance tracking for enhanced workplace security. Key skills: Computer vision, YOLOv9 implementation, safety analytics, compliance automation, and production model deployment.

  • YOLOv9
  • Computer Vision
  • Safety
  • Analytics
  • Compliance
Multi-AI Agents Platform (Telecom)

Multi-Agent / Telecom

Multi-AI Agents Platform (Telecom)

Read project details

Designed and prototyped a multi-AI agent platform for telecom operations, targeting zero-touch network automation. Successful PoC validated the agent topology on operator-scale workflows. Key skills: Multi-agent system design, telecom automation, system integration, agent orchestration.

  • Multi-Agent
  • Telecom
  • Automation
  • System Integration
  • Orchestration
  • Architecture
EasyOokla: Automated Network Analysis

Ookla / Data Analysis

EasyOokla: Automated Network Analysis

Read project details

Developed EasyOokla, a desktop application for automated analysis of Ookla speed test data. Streamlined data collection, analysis, and reporting for comprehensive network performance insights. Key skills: Data analysis, automation engineering, network performance analysis, Python development, and reporting systems.

  • Ookla
  • Data Analysis
  • Automation
  • Network Analysis
  • Python
  • Reporting

Many more AI and telecom projects — including proprietary work at Rakuten, Huawei, and earlier roles — aren’t publicly shareable. These give a representative sense of the problems I take on and how I ship.

About Ahmed

Product thinking.
Engineering depth.

AI Product and Innovation Lead with 16+ years in telecommunications. I spot real business problems, shape product vision and roadmap around them, and ship working AI systems — multi-agent platforms, RAG, and computer vision — end to end. Deep telecom context (RAN, OSS, CEM) makes the work sharper, not narrower.

Download my CV

Experience

Rakuten MobileAI and Innovation Lead for E2E Performance and CapacityOct 2024 — Present

Tokyo, Japan

  • Built and shipped AI Maestro — a multi-agent platform for automated KPI analysis, intelligent document search, and location-based insight visualizations. Saved 9,600 work-hours per year and cut root-cause-analysis time by ~80% (recognized with Rakuten Mobile Innovation Award and CEO Passionately Professional Award, Q3 2025).
  • Delivered an agentic RAG system over RMI E2E technical manuals, replacing manual lookup for performance and troubleshooting workflows.
  • Own the AI product roadmap for E2E network performance: use-case discovery, PoC-to-production progression, and prioritization against business outcomes.
  • Navigated privacy and security constraints by building on locally self-hosted LLMs; evaluated hundreds of models, frameworks, and tools for production fit.
  • Drive the innovation pipeline: multiple patent filings for novel AI applications in telecom (3 USPTO grants to date, more under review).
Rakuten SymphonyAI and Innovation Lead for Site ManagementAug 2023 — Oct 2024

Tokyo, Japan

  • Led cloud migration of Sympulse from monolith to microservices — delivered in 3 months.
  • Designed and prototyped a multi-AI agent platform for telecom operations, targeting zero-touch network automation.
  • Shipped a production LangGraph-based Advanced SQL Agent for natural-language querying of operational data.
  • Deployed YOLOv9/YOLOv10 computer-vision pipelines for site safety and compliance monitoring.
  • Ran the AI/ML use-case funnel for the mobility product line: discovery → PoC → demo → handover, with cross-functional product alignment.
Rakuten SymphonySenior Director, Sympulse Application SectionJan 2023 — Aug 2023
  • Owned product direction for the Sympulse application portfolio: market analysis, roadmap prioritization, and release planning.
  • Aligned engineering delivery with customer-facing outcomes across multiple product lines.
  • Set the groundwork for the AI-first product direction adopted in the subsequent role.
Rakuten SymphonyProduct Owner (Scrum)Jun 2022 — Dec 2022
  • Defined product vision and roadmap for OSS products; managed the backlog and release cadence.
  • Partnered with legal on patent development and infringement analysis — directly informed the patent portfolio work that followed.
  • Tracked technology trends and translated them into product opportunities the team could realistically deliver.
HuaweiCEM ExpertJan 2015 — Jun 2022

Riyadh, Saudi Arabia

  • Drove 4G/5G customer-experience analysis: Ookla throughput and latency, city-level KQIs/KPIs, and device benchmarking.
  • Built EasyOokla, a desktop application that automated speed-test data collection and reporting — the first working prototype of the automation thinking that runs through my AI work today.
  • Produced operator insights for major national events (Hajj, Ramadan, Riyadh Season) where network quality had direct business stakes.

My telecommunications career began in 2007 as a DriveTest Engineer in Egypt. I then worked on telecom installations, network optimization, and large-scale mobile infrastructure in Saudi Arabia. That hands-on foundation continues to inform my work in AI and innovation.

How I work

From the right problem
to a working product.

AI Product & Innovation

  • Product vision & roadmap
  • Use-case discovery from real business problems
  • AI product delivery cycles (PoC → pilot → production)
  • Cross-functional team leadership
  • Business-case & ROI framing
  • Telecom-AI domain depth

Generative AI & Multi-Agent Systems

  • LangGraph, CrewAI, AutoGen
  • Retrieval-Augmented Generation (RAG) systems
  • Local/self-hosted LLMs for privacy-sensitive deployments
  • Prompt engineering & evaluation
  • Multi-agent orchestration patterns
  • Agentic SQL and document-search agents

AI Engineering & Delivery

  • Python, FastAPI, production Python services
  • Computer vision (YOLOv9/v10) for safety & site ops
  • Cloud deployment & microservice migration
  • LLM/model evaluation for production fit
  • API & data-pipeline integration
  • Telecom data (KPIs, drive-test, OSS telemetry)

Recognition

A record of meaningful impact.

Passionately Professional Award — Rakuten Mobile CEO
Passionately Professional Award — Rakuten Mobile CEO — ceremony

Passionately Professional Award — Rakuten Mobile CEO

Read the story

Honored to receive the Quarterly Network Division 'Passionately Professional' Award from Sharad Sriwastawa, Co-CEO of Rakuten Mobile, in Q3 2025. This recognition reflects achievements in designing and leading AI Maestro, a multi-agent AI system integrating automated KPI analysis, intelligent document search, and innovative visualizations for location-based insights. Achieved 9,600 annual work hour savings and accelerated troubleshooting & Root Cause Analysis by ~80%, significantly enhancing productivity and delivering substantial business impact. Recognized for what can be achieved when strong principles meet trust, teamwork, and disciplined execution.

  • CEO Award
  • Multi-Agent AI
  • KPI Analysis
  • Productivity
  • Business Impact
  • Innovation
  • Leadership
View recognition
3 Granted USPTO Patents

3 Granted USPTO Patents

Read the story

USPTO Grant US 12308894 B2 — System, method, and medium for blockchain-enabled microwave antennas. USPTO Grant US 12328299 B2 — System, method and non-transitory computer-readable medium for an internet-enabled network radio node. USPTO Grant US 12587873 B2 — System, method, and medium for network drive testing via self-driving vehicle. Spans AI, blockchain, telecommunications, and autonomous-vehicle-based network testing. Additional applications are under review with the USPTO.

  • Patent
  • AI
  • Blockchain
  • Telecommunications
  • Autonomous Drive Test
  • USPTO
View recognition
2nd Place R-Pitch — ZenSpace AI

2nd Place R-Pitch — ZenSpace AI

Read the story

Achieved 2nd place at R-Pitch for ZenSpace AI, leading all aspects: financial planning, forecasting, pitch presentations, technical architecture, system design, AI product development, and market research. Recognized by Rakuten CEO.

  • Financial Planning
  • Forecasting
  • Pitching
  • Technical Architecture
  • System Design
  • AI Product Development
  • Market Research
View recognition
Certificate of Achievement for IP Excellence

Certificate of Achievement for IP Excellence

Read the story

Recognized by Rakuten's Patent League for outstanding innovation leadership.

  • Innovation
  • IP Excellence
  • Leadership
View recognition
Tokyo AI Speaker — Multi-AI Agents

Tokyo AI Speaker — Multi-AI Agents

Read the story

Delivered a keynote speech at TokyoAI on the future of Multi-AI Agents. Demonstrated skills in public speaking, applied AI engineering, and industry thought leadership.

  • Public Speaking
  • Applied AI Engineering
  • Multi-Agent
  • Thought Leadership
Rakuten Mobile Innovation Award

Rakuten Mobile Innovation Award

Read the story

Honored to receive the Rakuten Mobile Innovation Award for contributions in the E2E Performance and Capacity Department. Led the development of Phase 1 of AI Maestro, a pioneering product transforming E2E network performance using AI. Navigated significant privacy and security challenges by exclusively using locally self-hosted LLMs, rigorously testing hundreds of LLMs, frameworks, and tools to deliver real-world AI applications. Built an intelligent root cause analysis engine for critical core KPIs, demonstrating that innovation comes from challenging norms and building what doesn't yet exist.

  • Innovation
  • AI Product
  • Root Cause Analysis
  • Privacy
  • Security
  • Local LLM
  • Network Performance
A lifelong commitment to learning

Excellence has been a defining thread throughout my journey in life, beginning in elementary school where I consistently ranked among the top 5 students, earning numerous Academic Excellence Certificates and Good Conduct and Behavior Certificates—a pattern that continued through intermediate and secondary education, and throughout my college years, culminating in my graduation with a degree in Electrical and Computer Engineering from Egypt. This collection of achievements—carefully preserved in a folder I carry with me wherever I go—represents more than accolades; it embodies a lifelong commitment to continuous learning and excellence. Throughout my professional career, I have earned recognition from leading organizations including STC, and completed prestigious certifications such as IBM's Data Science Professional Certificate, PMP, Leadership Development programs, and NEBOSH Health & Safety credentials, among many others. Each certificate and award reflects not just achievement, but a dedication to mastery and professional growth that continues to drive my work in AI innovation today.

Let’s work together

Have a challenge in mind?

I’m open to high-impact roles, consulting engagements, and freelance projects.