t
talharazzaq194

Talha Razzaq

@talharazzaq194

AI Engineer, Voice AI and Workflow Automation Specialist

Pakistan
Anglais
Certaines informations sont présentées en anglais.
À propos de moi
I'm an AI Engineer building agentic systems, voice AI agents, and workflow automations for businesses that want to cut manual work and respond to customers faster. My background spans LangChain/LangGraph agent architecture, Deepgram + LiveKit voice pipelines, and n8n/Zapier integrations — backed by a Computer Engineering degree and hands-on production experience. I build systems that actually run, not demos, and I explain what I've built in plain language so you always know what you're paying for... Plus d’infos

Compétences

t
talharazzaq194
Talha Razzaq
hors ligne • 

Voir mes services

Mise en œuvre et déploiement de l’IA
I will build n8n automation, zapier workflow and api integration
Mise en œuvre et déploiement de l’IA
I will build ai voice agent, ai receptionist with vapi, retell and livekit

Expérience professionnelle

Z360

AI Engineer

Z360 • Temps plein

Aug 2025 - Present1 yr 1 mo

Built LLM evaluation and observability infrastructure in Python using LangFuse and LangChain, instrumenting multi-stage agent pipelines to trace individual model calls, isolate failure modes, and compare model and prompt variants against measurable latency and reliability baselines. • Led open-source development of Clave (clave.z360.biz), an AI-native project management platform combining issue tracking, real-time collaborative docs, and autonomous agents; owned repository structure, containerized development environments, and code review across contributors. • Engineered agentic workflows with LangGraph and BrowserUse that autonomously plan and execute multi-step tasks with retry and escalation logic, cutting manual intervention by 30%. • Designed production RAG pipelines (ingestion, chunking, embedding, retrieval) with LangChain and FAISS, exposed through Python REST services and deployed as containerized workloads. • Built low-latency real-time voice agent pipelines (STT → LLM → TTS) with Deepgram streaming ASR and LiveKit WebRTC, tuning end-to-end latency and turn detection for production deployments.