Resume / Curriculum Vitae

Sadman Sakib Rafi

AI and Machine Learning Engineer

Summary

AI and Machine Learning Engineer with hands-on experience in LLM-based systems, supervised fine-tuning, AI chatbot and AI-driven product development. Skilled in building and deploying real-world AI solutions using Python, FastAPI, AI tools, with a strong focus on scalable, production-ready systems.

Skills & Technical Stack

Professional Focus

Artificial IntelligenceMachine LearningSoftware Systems

Languages

PythonJavaScriptC++C#

Machine Learning & AI

Supervised Fine-TuningClassificationNLPLLM ArchitecturesPrompt EngineeringLoRA

Frameworks & Libraries

Hugging FaceScikit-learnLangChainCrewAINext.js

AI Tools & Systems

OllamaGeminiMCPVector EmbeddingsPinecone

Backend APIs

FastAPINestJSASP.NETREST API Design

Portfolio Systems

Multi-Agent AI Assistant System

Backend-focused AI assistant orchestration platform using FastAPI and multi-agent architecture. Used agent routing, short-term memory handling, vector embeddings, and LLM abstractions for scalable assistant reasoning and responses.

FastAPIMulti-Agent ArchitectureVector EmbeddingsLLM Abstractions

HireSenseAI - CV Ranking System

Repository

AI-driven recruitment platform that analyzes and ranks resumes against job requirements. Supports batch CV ingestion, deep semantic reasoning, gap detection, and explainable decision scoring. Built with a hybrid LLM pipeline, Pinecone similarity search, and local embedding models.

GeminiOllamaPineconeLocal Embeddings

MeetingMind

AI meeting intelligence pipeline using a multi-tool MCP server with a unified LiteLLM client. Extracts action items with owner and deadline assignment, detects blockers and sentiment, scores meeting quality, and syncs tasks to Notion and Trello with cross-session memory.

MCP ServerLiteLLMNotionTrello

AI Tutorial Crew

Multi-agent system using CrewAI and Ollama-Mistral that generates tutorials based on user input. Supports local inference for privacy and speed.

CrewAIOllamaMistralLocal Inference

Automated Conversation Platform

Real-time AI assistant platform for automated business interactions. Designed backend orchestration, context memory, tool calling, and LLM-driven response handling with extensible APIs for voice and chat.

Context MemoryTool CallingLLM ResponsesVoice & Chat APIs

Machine Learning Projects

Supervised Fine-Tuning for Lead Qualification

Built a supervised fine-tuning pipeline using Qwen3.5-0.8B to convert unstructured sales leads into structured data.

Qwen3.5-0.8BSupervised Fine-TuningStructured Data

LoRA Fine-Tuning (NLP)

Fine-tuned DistilRoBERTa on the IMDB dataset using LoRA for efficient sentiment classification.

DistilRoBERTaLoRAIMDB DatasetSentiment Classification