Zeeshan K.

Agentic Software Engineer (Open AI Codex, Google AntiGravity, Claude Codex) 23 years of Software Development experience. Member of UpWork since 2010. Data Science & Analytics | AI & Machine Learning | Machine Learning Generative AI Modeling | Deep Learning | AI Data Annotation & Labeling | Knowledge Representation Data Analysis & Testing | Data Extraction/ETL Data Analytics | Data Visualization | Experimentation & Testing | Data Extraction | Data Processing I help startups and growing businesses build production-ready Generative AI systems — with a strong focus on AI Voice Agents and advanced, multi-source RAG architectures that go far beyond simple demos. I architect and deploy production-grade custom AI systems that unify data, automate workflows, and integrate intelligence directly into business operations. Whether through machine learning models, LLM-powered agents, or API-driven automation pipelines, I build scalable solutions designed for accuracy, maintainability, and real-world deployment — not experimental demos. Machine Learning | Deep Learning | Generative AI Modeling | AI Integration | Scripting & Automation | Back-End Development | Data Analytics | AI Chatbot Development | Natural Language Processing | Artificial Intelligence | AI Agent Development ✅ What I actually build: 🔹 AI Voice Agents • Voice agents using Vapi AI and Twilio for calling workflows • STT with AssemblyAI / Deepgram • LangGraph-based agents using Anthropic, OpenAI, and Groq models • Tool-calling voice agents for real-time actions and external data access • TTS with ElevenLabs / Cartesia (low-latency, natural voice) • Reliable conversational flows designed for production use 🔹 Agentic RAG Systems • Multi-source retrieval (PDF, CSV, SQL, Web) • Vector databases: FAISS, Qdrant, ChromaDB • Context assembly, hybrid retrieval, reranking, and validation to reduce hallucinations 🔹 MCP-Style AI Architectures • Modular servers for ingestion, retrieval, reasoning, and tools • Clean separation of data sources, agents, and LLM logic 🔹 Automation & Data Pipelines • Python automation, API integrations, and n8n workflows • Data extraction, scraping, and normalization (Playwright, web sources) 🛠 Tech Stack: Python, LangChain, FAISS, Qdrant, Chroma, Vector DBs, LLMs (OpenAI GPT, Anthropic Claude, LLaMA, Gemini, Groq, Ollama local model), Pandas, SQL, Flask, FastAPI, Streamlit, APIs, n8n, Playwright, Docker, VS Code, Scikit-Learn, PyTorch, TensorFlow I care deeply about production reliability, hallucination control, and delivering AI systems that work end-to-end. I also handle deployment, infrastructure setup, and production operations, so clients get a complete end-to-end AI system. 📩 If you need an AI solution designed, implemented, and deployed in production, let’s talk.