t
tahaahmad15

Taha Ahmad

@tahaahmad15

AI System Engineer

Pakistan
Englisch, Französisch, Chinesisch
Einige Informationen werden in englischer Sprache angezeigt.
Über mich
I'm an AI/ML Engineer with hands-on experience delivering production-ready AI systems that actually move business metrics. My expertise includes: RAG & LLM Systems — Retrieval-Augmented Generation, agentic workflows, tool-calling, fine-tuning (SFT/PEFT) Voice AI — real-time STT/TTS pipelines, conversational voice agents Computer Vision: YOLO, SAM, ByteTrack, video annotation pipelines, model benchmarking Backend & APIs: FastAPI, Python, PostgreSQL, Docker, Redis, Google Cloud Previous roles, results 60% reduction in infrastructure costs via distributed inference systems... Mehr lesen

Kompetenzen

t
tahaahmad15
Taha Ahmad
offline • 
Durchschnittliche Antwortzeit: 2 Stunden

Meine Dienstleistungen

KI-Websites & -Software
I will build a custom rag chatbot using langchain, llamaindex, and openai
Ab150 $ / ProjektWeitere Details

Portfolio

Arbeitserfahrung

SymBULL

AI/ML Engineer

SymBULL • Vollzeit

Sep 2025 - Present • 10 mos

Built a distributed inference system using Docker and Redis queues, reducing infrastructure costs by 60% through on-demand GPU utilization. Automated video annotation pipelines using YOLO, eliminating 5 annotators and saving $24K annually. Extended Label Studio into an enterprise-grade annotation platform with RBAC and GCP storage integration, supporting 50+ annotators across 10TB+ of data while saving $10K/year in licensing. Led benchmarking of vision models including SAM, YOLOv8/v10, and Grounding DINO for real-time deployment. Built a scraping and pricing pipeline processing 10K+ products. Led full-stack development of the AI Empath MVP using React, FastAPI, and PostgreSQL.

Software Engineer

Aspire365 • Vollzeit

Jul 2023 - Jun 2025 • 1 yr 11 mos

Built an end-to-end agentic AI financial reconciliation system using RAG and tool-augmented agents, reducing reconciliation time by 80% with 100% accuracy. Developed LLM-powered conversational and voice agents (STT/TTS), reducing support workload by 60% and operational costs by 50%. Automated business workflows using browser automation and Google Apps Script, cutting repetitive effort by 60%. Architected backend systems for a multi-entity ERP platform covering POS, inventory, HR, and accounting — supporting 1K+ daily transactions and reducing accounting overhead by 35%.