
Ata Jafari
AI Software Engineer
Habilidades

Conheça meus serviços

Experiência profissional
Freelance
Período integral • 2 yrs 2 mos
AI Software Engineer
Dec 2024 - Aug 2026 • 1 yr 8 mos
AI Software Engineer Veneto, Italy Worked as an AI Software Engineer for hift, where I worked on couple of projects: *Themis– SaaS platform for Italian law firms with real-time collaboration and AI-assisted legal work flows. • Developed an AI legal assistant with streaming responses, dynamic memory, and retrieval of relevant legal context using LangChain. • Built an evaluation and observability framework using MLflow and LangSmith. • Designed the application architecture and database schema. • Developed an agentic RAG workflow to support review of legal documents against retrieved Italian legal sources and contractual rules. • Deployed application infrastructure on AWS using Terraform. • Stack: LangGraph, PostgreSQL, Redis, Docker, Express.js, Knex.js. *Byse– AI-powered product recommendation application using Amazon marketplace data. • Built a multi-agent recommendation workflow using LangChain to interpret user intent and generate structured product recommendations. • Developed the FastAPI backend and APIs supporting the recommendation system. • Deployed the backend using Terraform and implemented automated testing and CI/CD workflows with GitHub Actions. • Stack: Python, FastAPI, LangChain, MongoDB, Docker, Terraform, GitHub Actions. *AiStudios– Automated marketing creative generation system. • Designed and deployed a background-removal microservice to prepare scraped images for creative generation workflows. • Built scraping pipelines to extract structured text, metadata, and media assets from websites and YouTube videos. • Implemented agent-based creative generation workflows using predefined campaign templates. • Stack: Python, FastAPI, LangChain, Docker, AWS, Terraform, GitHub Actions.
AI Engineer & Consultant
Jun 2024 - Dec 2024 • 6 mos
Restaurant Club– Website generation system using restaurant-owner data and AI-generated content. • Built a website-generation backend using FastAPI, MongoDB, and DSPy to generate structured, multi-section content from restaurant-owner data. • Integrated GrapesJS templates with generated content to assemble restaurant websites. • Stack: Python, FastAPI, MongoDB, DSPy, Docker, GrapesJS. Copywriter Agent– AI agent for generating Amazon product descriptions. • Built a CrewAI workflow with specialised agents for product research, copywriting, and content optimisation. • Integrated MongoDB for product data management and template storage. • Stack: Python, CrewAI, MongoDB, Docker