Adam MABROUK

- [Nexy.works] Developing a Graph-RAG solution to recommend similar profiles based on discussed topics and interests (ETL + GenAI). - [UnReal Estate] Developing a chatbot for real estate client qualification, with integrated loan simulation. Link : https://unrealestate.fr/ - [Amano Restaurant] Developing a digital waiter assistant using VAPI + N8N that takes orders, performs upsells, and sends them to a Telegram group. Link : https://serving-restaurant-assistant.web.app/ - [Internal Project] Designed a voice assistant connected to web services to automate appointment scheduling and streamline client engagement. Link : https://www.youtube.com/watch?v=Q2WgB7EMS84 - [ABIBOO Studio] Replatformed a real estate application to AWS AppStream to enable cloud-native delivery and scalability. Link : https://www.youtube.com/shorts/f-2da9WNylY Tech: Terraform, AWS, IAM, Python, AWS AppStream, Databricks, WebRTC, GCP, Dialogflow CX

- Led the strategic definition and implementation of a unified service catalog encompassing CloudOps, SysOps, SecOps, and DataOps, collaborating directly with Veolia's Information Systems Direction at St Maurice. - Engineered the catalog for a transverse infrastructure team, standardizing the delivery of packaged services (primarily leveraging AWS and GCP) to diverse internal business units (e.g., Consumers, Operations). - Drove the alignment of service offerings with internal demands, ensuring seamless integration of the service catalog with the IT CMDB. - Established robust governance by incorporating transverse FinOps and Security teams as key approvers within the new catalog's operational workflows. - Successfully replicated and adapted this service catalog framework for the 'Industrial' Business Unit at the IT/OT industrial frontier, addressing specific DMZ services (e.g., Bastion, Messaging Gateways, PCU Servers for telemaintenance) and tailoring the approach for regional branches. - Technologies: AWS, GCP, ServiceNow, CMDB, ITIL/ITSM Frameworks, Cloud Security Principles, DataOps Principles, Service Catalog Management, Service Now

- Developed RESTful APIs using Python and Lambda powertools, implementing a fully AWS Serverless stack for enhanced backend functionality. - Optimized video processing and inference time, achieving a 50% reduction (from 2 minutes per video to 1 minute per any video batch) by scaling the prediction microservice horizontally. - Spearheaded FinOps and MLOps efforts, resulting in a 72% reduction in cloud costs by reworking AI models inference code, migrating the stack to a serverless architecture, and decommissioning redundant infrastructure (Zombie IT). - Implemented robust DevOps practices, integrating CI/CD pipelines using Cloudformation, AWS SAM, Docker, and GitHub Actions for smooth and efficient code deployment. - Designed network architecture and integrated the ImVitro SAS platform with Timelapse Incubators Rest APIs (Embryoscope, Geri, Miri), automating embryo video extraction and evaluation. - Created and implemented a data anonymization ETL solution using Glue, facilitating internal dashboarding and accurate KPI measurement. Technologies used : * AWS : IAM, API Gateway, EC2, Cloudwatch, Lambda, ECS Fargate, ECR, DynamoDB, Glue * Python: Pytest - Unit Testing, Lambda Powertools -Microframework * DevOps: Github, Github Actions, Docker, Cloudformation & SAM- IaaC

- Backend development of a SAS tool for cost control on several Cloud platforms, using an AWS Serverless Stack, all embedded in a documented REST API. - Participation in Brainstorming, writing of User Stories, of the different features. - Implementation of a CI / CD chain on Gitlab ( Build -> Unit Test -> Terraform Deploy ). - Development of an internal framework allowing the automation of the financial audit of a client cloud platform - Piloting FinOps missions ( Organizing Workshops, Launch of audits, Restitution, support on implementations ). - Participation in the recruitment process Technologies used : * AWS : IAM, API Gateway, EC2, Cloudwatch, Lambda, ECS Fargate, ECR, RDS - PostgreSQL, EMR - Spark * Python: Poetry - Dependency Management, Peewee - ORM, Marshmallow - Serialization, Pandas, Pytest - Unit Testing, Sklearn, Prophet * DevOps: GitLab, Docker, Terraform

-[NLP] In-house development of NLP methods for the detection of Covid-19 risk factors in scientific articles. -[NLP] Implementation of a Twitter observatory to monitor the opinion of the population in India / South Africa, concerning the topic of vaccination. -[NLP] Implementation of a pipeline for information extraction from PDF files (text and images) and referencing in structured format. -[Cloud] Implementation of a single-tenant architecture to migrate the Visual Studio Code IDE to the Azure public cloud and access it from the browser. -[Cloud] Participation in the migration project of Danone's old information system to the AWS cloud platform ( Availability zones, VPC, Direct Connect, EC2 ).

Skills: Communication · Linux · Mathematics

Skills: Communication · Linux

Skills: Communication · Linux

Skills: Communication · Linux

Skills: Communication