You will get a Custom AI Agent with custom system prompts and optional vectorDB for RAG
Project details
I build custom AI agents with system prompts engineered for your real workflows—not generic chat templates.
Starter focuses on a dependable conversational agent on agreed topics and scenarios, with optional tools only where we explicitly scope them. Standard adds retrieval from one primary knowledge source you provide (for example docs, a wiki, or a bounded file corpus), including ingestion and grounded answers. Advanced is a production-minded build: multi-source RAG, stronger evaluation and retrieval tuning, and optional MCP-style integrations aligned to your environment.
Each tier includes documentation and source code as defined in the project. Before build work begins, we confirm model provider, hosting, data access, security constraints, and acceptance criteria so delivery stays predictable.
Clients work with me for crisp scope, honest limits on what RAG can guarantee, scenario-based acceptance, and a handoff you can operate and extend.
Starter focuses on a dependable conversational agent on agreed topics and scenarios, with optional tools only where we explicitly scope them. Standard adds retrieval from one primary knowledge source you provide (for example docs, a wiki, or a bounded file corpus), including ingestion and grounded answers. Advanced is a production-minded build: multi-source RAG, stronger evaluation and retrieval tuning, and optional MCP-style integrations aligned to your environment.
Each tier includes documentation and source code as defined in the project. Before build work begins, we confirm model provider, hosting, data access, security constraints, and acceptance criteria so delivery stays predictable.
Clients work with me for crisp scope, honest limits on what RAG can guarantee, scenario-based acceptance, and a handoff you can operate and extend.
Machine Learning Tools
Azure Machine LearningWhat's included
| Service Tiers |
Starter
$3,000
|
Standard
$5,900
|
Advanced
$14,500
|
|---|---|---|---|
| Delivery Time | 12 days | 21 days | 60 days |
Number of Revisions | 2 | 2 | 3 |
Model Validation/Testing | - | - | - |
Model Documentation | |||
Data Source Connectivity | |||
Source Code |
Frequently asked questions
5 reviews
(5)
(0)
(0)
(0)
(0)
This project doesn't have any reviews.
PA
Patrick A.
Dec 22, 2021
Build pipeline using Java for Azure DevOps
GR
Geoffrey R.
Dec 9, 2021
Azure DevOps k8s deployment pipline
Well worth the rate. Very knowledgeable and self-sufficient
GR
Geoffrey R.
Jun 16, 2020
Jenkins AKS pipeline
Great ability to grasp the requirements and provide a solution with minimal supervision!
Their extensive prior experience allows them to re-use CICD code and work very efficiently
Their extensive prior experience allows them to re-use CICD code and work very efficiently
IT
Irvin T.
Dec 22, 2018
Write Chef cookbook for nodejs
Daeda Technologies was able to re-use scripts from their own repositories to complete this job very quickly! The slightly higher rate is entirely worth it.
IT
Irvin T.
Dec 21, 2018
Write Chef cookbook for mongodb cluster and provide ongoing guidance
Van Truong (Daeda Technologies) used an existing repository of reference implementations to complete this work very quickly and efficiently. They were very professional, self-sufficient and clearly subject matter experts in DevOps and Chef.
About Van
AI Solutions Architect
Woodbridge, United States - 4:17 pm local time
I’m an IT Consultant and AI Solutions Architect with 17+ years of experience building and scaling production systems across cloud, DevSecOps, software development, and AI.
I specialize in stepping into complex or underperforming environments, identifying root causes, and delivering systems that are scalable, secure, and maintainable long-term.
I’ve worked with large enterprises including PricewaterhouseCoopers, where I helped scale a DevOps organization from a small team into 100+ engineers across a multi-team operating model—establishing CI/CD pipelines, standards, and governance to support enterprise delivery at scale.
AI & Modern Systems (Recent Work)
My recent work is focused on production-grade AI systems, not just prototypes:
RAGline — A modular pipeline for AI/RAG workflows
• Document + media ingestion (PDF, Office, audio)
• Transcription (Whisper.cpp), speaker diarization (pyannote)
• Structured output for retrieval and agent workflows
• Containerized, local-first where possible, enterprise-ready
Family AI Platform
• Multi-tenant architecture (channels, groups, real-time interaction)
• AI embedded directly into user workflows
• Token-based usage and context-aware responses
• Combines structured long-term memory with conversational UX
Document Intelligence / OCR Systems
• PDF + image ingestion pipelines
• Hybrid extraction (native text + OCR / vision models)
• LLM-driven structured outputs (strict JSON schemas)
• Designed for reliability, validation, and downstream UI consumption
Core Consulting Capabilities
• DevSecOps architecture (CI/CD, security integration, release automation)
• Cloud architecture (AWS, Azure, microservices, containerization)
• AI system design (RAG pipelines, LLM integration, data pipelines)
• System scaling (performance, cost optimization, reliability)
• Technical leadership (team structure, standards, delivery alignment)
I operate both hands-on and as a technical lead. Depending on scope, I can:
• Execute independently
• Use AI to accelerate delivery
• Bring in a small, trusted team for larger engagements
Selected Technologies
AWS, Azure, Terraform, Jenkins, Azure DevOps, Docker, Linux
RAG pipelines, LLM APIs, Whisper.cpp, pyannote, OCR/document processing
Node.js, PHP, SQL/NoSQL (PostgreSQL, MongoDB, MySQL, Oracle)
More information: www\DaedaTechnologies\com
Steps for completing your project
After purchasing the project, send requirements so Van can start the project.
Delivery time starts when Van receives requirements from you.
Van works on your project following the steps below.
Revisions may occur after the delivery date.
Kickoff Call
confirm use cases, tone, languages, model/provider, hosting, and data access; define acceptance scenarios.
Prompt Architecture
I deliver prompt architecture, safety/behavior notes, and (if applicable) retrieval design; knowledge transfer to client to enable them to modify prompts as needed.