Saif

AI Engineer

mojosolo • Tiempo completo

Aug 2022 - Present • 4 yrs 1 mo

• Designed and deployed production LLM systems using LangGraph, LangChain, OpenAI and Anthropic SDKs, including RAG pipelines, tool-calling agents, structured-output workflows, and real-time AI processing services. • Built agentic workflows that processed live transcripts, invoked external tools, applied predefined business logic, and generated structured feedback for downstream applications. • Developed Model Context Protocol integrations and reusable tool interfaces that enabled AI agents to interact with external services, internal APIs, and application data. • Built and optimized retrieval systems using Pinecone, PostgreSQL, and pgvector, improving response relevance through chunking strategies, metadata filtering, retrieval tuning, and context management. • Implemented LangSmith and Langfuse for tracing, prompt monitoring, workflow debugging, and analysis of agent and retrieval behaviour. • Developed AI-assisted software delivery workflows that processed feature requests and GitHub issues, generated code changes, ran automated test suites, and opened pull requests for review. • Architected and maintained 5 independently scalable microservices in Python and Laravel, with centralized OAuth2 and JWT authentication across multiple business domains. • Containerized applications using Docker and Docker Compose and managed deployments across AWS EC2, S3, Vercel, Linux, and PM2-based environments. • Launched and maintained 3 customer-facing applications with 99.9% uptime while reducing average load time from 2.5 seconds to 1.2 seconds. • Automated internal workflows that saved more than 200 hours annually and reduced manual reporting effort by 75%. • Mentored 2 junior developers in Python, LangChain, API development, and code-review practices, improving code-review pass rates by 40%