Muhammad Daniyal A.

Most AI projects stall between the demo and daily use. I build the part that gets them over that line: AI agents, voice agents and automations that run reliably inside the tools your team already uses, with logging, safe retries and a person approving anything that matters. I'm a senior engineer with ten years of shipping production software. I build with Claude, OpenAI, n8n, Make, Twilio and LiveKit, plus Next.js, Node.js and Python when a workflow needs real code. RECENT WORK Recruitment AI agents (n8n, Apollo, Claude, Bullhorn): agents that watch target firms, score buying and hiring signals, write tailored outreach and add each lead to the CRM with a drafted message for a consultant to approve. A consultant's daily research went from 3 hours to a 5 minute morning review. A second agent matches vetted candidates with employers within a 30 minute commute and writes a pitch for each match. German outbound voice agent (LiveKit, Twilio, Claude Code): took over a Retell and Make build that struggled on real calls and rebuilt it on LiveKit with Twilio, fixing turn taking, interruptions and latency, and building the AI disclosure and opt-out handling into every call. Email outreach platform (Next.js, Node.js, MySQL): built for a US construction estimating company, reaching 93% inbox placement and cutting their marketing cost in half. 5 star reviews for OpenAI integration and RAG optimization, for recruitment AI automation, and for an n8n and AI build at a Canadian startup that hired me again. WHAT I CAN DO FOR YOU Finish or fix half-built AI agents, Claude setups and automations Build AI agents with tool calling, approval steps and audit logs Set up Claude Code and custom Claude skills for your team Build AI voice agents and phone receptionists on Twilio, LiveKit or Retell Connect your CRM, inbox, calendar and sheets with n8n, Make, Zapier or custom APIs Review AI-generated code and architecture before it reaches production HOW I WORK You get a small first milestone you can test within days, short written updates, and documentation your team can run without me. I use AI coding tools to move fast, but the design, review and testing stay with me, because I know where AI breaks. Tell me the task that eats the most hours of your team's week. I'll tell you honestly whether an agent is the right fix and what a first version would take.