Muhammad Ali Abbas

Lead Machine Learning Engineer at Idrak Ai Ltd.

I am Muhammad Ali Abbas (Ali), a dedicated Machine Learning Engineer and researcher with expertise in Speech AI, NLP, and Computer Vision. Over the years, I have made significant contributions to AI research, publishing papers on diagnosing COVID-19 using Swin Transformer, predictive maintenance in motor faults, and pneumonia detection through hybrid deep learning methods. My research also includes AI-based histopathological diagnosis of lung and colon cancers and identifying health factors during pandemics using feature selection and neural networks. These works reflect my commitment to solving real-world problems through AI innovation. Currently, as a Lead Machine Learning Engineer at Idrak AI Ltd., I have developed cutting-edge solutions such as outbound and inbound calling bots, CFO agents for financial analysis, and generative AI agents for various domains. Leveraging technologies like PyTorch, Huggingface, LangChain, LangGraph, and LlamaIndex, I have optimized TTS and ASR systems, created multilingual TTS systems, and engineered noise cancellation tools for real-time applications. My work also includes designing analytics dashboards with speaker diarization and text retrieval, demonstrating my ability to build scalable, impactful AI-driven products. I am actively seeking Ph.D. opportunities in Speech AI, NLP, or Computer Vision, focusing on generative models, multimodal AI, and self-supervised learning. Additionally, I am open to Machine Learning roles and collaborations with investors to scale innovative AI products. With my strong foundation in research and development, along with expertise in building advanced conversational systems and generative AI agents, I aim to push the boundaries of AI and contribute to transformative advancements in the field.

Educations

Experiences

Publications

M. Ali Abbas, S. Edrees, M. Maryam and Others

A Novel-based Swin Transfer-Based Diagnosis of COVID-19 Patients

Intelligent Automation & Soft Computing, Volume 29, January 2022, DOI: 10.32604/iasc.2023.025580

M. Ali Abbas, S. Saeed, M. Maryam and Others

A Turf Feature Selection-Based Technique for Predicting Factors Affecting Human Health During Pandemic

Life, MDPI, Volume 12, January 2022, Pages 1367, DOI: 10.3390/life12091367

Posts

From LLMs to LightRAG: A New Era of Smarter Retrieval

Dive into the evolution from traditional LLM-based RAG systems to LightRAG, an innovative approach for retrieval-augmented generation that emphasizes efficiency and smarter context retrieval.

From Local to Global: The Story of Graph RAGs

Explore the evolution of Retrieval-Augmented Generation (RAG) systems to Graph RAGs, which enhance global context retrieval through graph-based knowledge representations. This article narrates their journey from local retrieval systems to globally connected insights.

Life Events

2025

  • Check

    Developed portfolio website

    Created a personal portfolio website to showcase my projects, research, and professional achievements in AI and machine learning.

  • Check

    Completed CFO-II Conversational Agent for Financial Analysis

    Successfully developed CFO-II, a multimodal voice-to-voice conversational agent designed for advanced financial analysis.

2024

  • Check

    Earned multiple AI certifications

    Achieved certifications in Retrieval Augmented Generation for Production with LangChain & LlamaIndex by Activeloop, Training and Fine-tuning LLMs for Production by Activeloop, LlamaIndex and Intel, and LangGraph by LangChain Academy.

  • Check

    Started developing CFO-II for financial analysis

    Began the development of CFO-II, a multimodal voice-to-voice financial analysis agent, in late December.

2023

  • Check

    Joined GC Faisalabad for B.Ed

    Pursued a Bachelor of Education (B.Ed) to focus on leveraging AI for enhancing student learning and advancing educational experiences.

  • Check

    Promoted to Lead Machine Learning Engineer at Idrak AI Ltd.

    Took on a leadership role to oversee a team of 7 ML engineers and drive scalable AI projects, including generative agents and advanced ASR/TTS systems.