Browse categories
Explore
Fiverr Pro
English
$
USD
I will build custom ai agents, rag systems, langgraph and crewai automation
YS Malhara
About this gig
I specialize in building production-grade AI Agents and Retrieval-Augmented Generation (RAG) applications that can reason, use external tools, collaborate in multi-agent teams, and dynamically query your private knowledge bases. I turn raw LLMs into reliable enterprise systems.
What I Can Build For You:
- Stateful Multi-Agent Systems: Complex, deterministic agent workflows using state machines.Perfect for scenarios needing conditional loops, error self-correction, and Human-in-the-Loop (HITL) approval steps.
- Advanced RAG Pipelines: High-accuracy retrieval networks featuring hybrid search, parent-document retrieval, reranking, and Graph RAG to ensure zero hallucination on your proprietary data.
- Custom Tool & API Integration: Equipping agents with the ability to safely interact with your databases, CRM, Slack, Gmail, or any custom REST API.
Why Choose My Architecture? I don't build brittle prototypes that loop endlessly and waste your API tokens. Every agentic system I deploy features strict state boundaries, explicit token-consumption caps, and native observability so you can audit exactly why an agent made a specific decision.
Please message me before placing an order!
Get to know YS Malhara
YS Malhara
AI and Machine Learning Engineer
- FromSri Lanka
- Member sinceOct 2020
- Avg. response time1 hour
- Last delivery1 year
Languages
English
I am an AI/ML Engineer specializing in Machine Learning, Deep Learning, Generative AI, and Agentic AI systems. I develop intelligent solutions that transform data into actionable insights and automate complex workflows.
I have experience building AI solutions for malware detection, financial forecasting, intelligent assistants, document analysis, recommendation systems, and autonomous decision-making systems.
Whether you need an AI-powered application, custom machine learning model, Agentic AI workflow, or end-to-end AI system, I can help bring your idea to reality.
FAQ
What is the difference between CrewAI and LangGraph, and which do I need?
CrewAI is incredible for rapid prototyping and workflows that mimic a team of human specialists. LangGraph is a lower-level, graph-based state machine used for enterprise systems that require strict control, complex conditional loops, error recovery, and human intervention points.
How do you prevent agents from "looping" infinitely and driving up API costs?
I build strict guardrails into every system. For CrewAI, we implement hard task termination conditions. For LangGraph, we utilize explicit node structures with max iteration limits and integrate LangSmith to track token costs and catch anomalous behavior immediately.
Can the AI agent ask for my permission before taking a critical action?
Yes! Using LangGraph's native state persistence and checkpointing, we can set up "Human-in-the-Loop" gates. The agent will pause, save its exact state, notify you (via Slack, email, or a UI) to review its work or tool call, and resume cleanly once you hit approve.
Will the application work with my company's internal data?
Absolutely. We will build a customized RAG (Retrieval-Augmented Generation) pipeline that securely ingests your PDFs, database schemas, or API documentation, ensuring the agents only generate answers grounded directly in your factual data.
