Upwork home
Sign up
Project Catalog
DesignDesign
Video & AudioVideo & Audio
Development & ITDevelopment & IT
MarketingMarketing
Writing & TranslationWriting & Translation
MoreMore

You will get Legacy Application Modernization with Model Context Protocol (MCP)

Oguz K.Status: Offline
Oguz K.
4.7
Top Rated
Preview image

Let a pro handle the details

Buy Web Application Programming services from Oguz, priced and ready to go.
Oguz K.Status: Offline
Oguz K.
4.7
Top Rated
Preview image

Let a pro handle the details

Buy Web Application Programming services from Oguz, priced and ready to go.

Project details

This project offers a unique solution for modernizing legacy applications without disrupting existing systems. By implementing Model Context Protocol (MCP), I create a smart, intuitive layer that interacts seamlessly with your legacy APIs, WCF services, or databases—transforming them into modern, user-friendly interfaces using React, Angular, Blazor, or Razor Pages.

What sets this project apart is its focus on user experience and adaptability. Instead of rewriting the entire system, I build a bridge that simplifies user interaction, accelerates employee onboarding, and extends the life and usability of your legacy software—fast, cost-effective, and maintainable.
Programming Languages
Python, TypeScript, C#
Coding Expertise
Cross Browser & Device Compatibility, Performance Optimization, Security
What's included
Service Tiers Starter
$750
Standard
$1,500
Advanced
$3,000
Delivery Time 7 days 14 days 21 days
Number of Revisions
123
Number of Pages
3610
Design Customization
Content Upload
-
Responsive Design
Source Code
Optional add-ons You can add these on the next page.
Fast Delivery
+$750 - $3,000
Additional Revision
+$250

Frequently asked questions

4.7
11 reviews
82% Complete
(9)
9% Complete
(1)
9% Complete
(1)
1% Complete
(0)
1% Complete
(0)

CT

Carlos T.
5.00
Jun 6, 2026
Senior Voice AI Architect Needed for Independent Technical Review— 5-10 hours Excellent independent review. Oguz delivered a thorough, evidence-based assessment, validated the architecture where sound, and caught real issues my developer missed — including a serious correctness bug. Professional, on time, scope respected. Would hire again.

ND

Niels D.
5.00
May 6, 2026
NET 8 REST API for server-side SVG to PNG rendering with watermark and visual enhancement

BL

Bobby L.
3.00
Mar 24, 2026
🚀 Senior .NET Developer

DL

David L.
5.00
Mar 10, 2026
Python developer (with some AI experience) -- OÄźuz OÄźuz is a fantastic developer with an exceptional command of various AI tools and a breadth of knowledge. I would highly recommend him to anyone looking for a developer.

DL

David L.
5.00
Sep 24, 2025
Python developer (with some AI experience)
Oguz K.Status: Offline

About Oguz

Oguz K.Status: Offline
Senior AI Engineer | Claude Voice AI RAG LangGraph Python .NET
95% Job Success
4.7  (11 reviews)
Istanbul, Turkey - 4:16 pm local time
I build AI systems that ship to production — with evaluation pipelines, human-in-the-loop oversight, and governance built in. Claude agents, voice AI, RAG, multi-agent orchestration — documented publicly so you can see how I work before you hire me.

20+ years of professional engineering. 6,400+ hours on Upwork. 100% JSS. Top Rated Plus.

âś… What I Build

🤖 Claude Agent SDK & MCP Servers
Design and ship agentic Claude workflows: multi-step tool use, MCP server
configuration, CLAUDE.md context management, Claude Code hooks, and production
multi-agent delegation. Full series published — see portfolio.

🎙️ Voice AI Engineering
Sub-200ms latency voice agents using ElevenLabs bidirectional WebSocket TTS + Twilio
ConversationRelay/Media Streams. LangGraph barge-in, Redis call checkpointing,
escalation routing. Built across a 4-episode published series — see portfolio.

🧠 RAG Pipelines & Hybrid Search
Production RAG with Qdrant hybrid search (BM25 + dense, RRF fusion, cross-encoder
reranking), ChromaDB, pgvector. NLP-to-SQL with Vanna.ai. 3-episode published
series — see portfolio.

🕸️ Multi-Agent Orchestration
LangGraph Supervisor pattern — multi-model councils (Claude + GPT-4 + Gemini),
bounded agent authority, checkpointed state, HITL escalation paths.

đź“Š AI-Enhanced Reporting & BI
Vanna.ai NLP-to-SQL, interactive dashboards, token-budgeted Q&A on live data.
Deployed at a media distribution platform (enterprise / private).

🏗️ Legacy Modernization + AI Layer
.NET Framework → .NET 10 migration with clean architecture, then AI capabilities
layered on the modernized foundation. Silverlight → WebAssembly via OpenSilver.

đź”— Enterprise & Full-Stack Engineering
- Python (FastAPI / Django), .NET (C# / ASP.NET Core), and Node.js (TypeScript).
- Twilio (Voice/SMS), Firebase, AWS (Cognito, S3), RabbitMQ, payment gateways.
- High-performance bridges for cross-platform UI (OpenSilver / Blazor / DevExpress).


🏆 Why Clients Hire Me

- 100% Job Success | Top Rated Plus.
- Everything I build is documented publicly — code and video series available in
the portfolio section below.
- Production precedents in every lane: Claude agents, Voice AI, RAG, multi-agent,
legacy modernization.
- Full-stack depth: vector DB and LLM backend through to frontend chat UI.
- Governance-first approach: bounded agent authority, HITL escalation paths,
evaluation pipelines, and audit-ready logging — not bolted on, designed in.

đź“Ś Recent Work

- ai-decision-room (public, contract closed): Multi-model council agent — Claude, GPT-4,
and Gemini each reason independently; a LangGraph moderator synthesizes decisions.
Delivered within a 2-week sprint.

- A media distribution platform: Claude API + RAG pipeline for NLP-to-SQL on a
media supply chain. FastAPI, ChromaDB, Playwright test suite. Production.

- Voice AI series (4 episodes, published): ElevenLabs sub-200ms TTS, Twilio bridge,
multi-modal voice design, LangGraph barge-in + Redis checkpointing.

- Qdrant Hybrid Search series (3 episodes): BM25+dense hybrid, RRF fusion,
FastAPI + cross-encoder reranking. Production RAG pipeline architecture.

🛠️ Stack & Skills

• AI / LLM — Claude API, Claude Agent SDK, MCP Servers, OpenAI, LangGraph, LangChain, RAG, Prompt Engineering, AI Agents
• Voice AI — ElevenLabs, Twilio (Media Streams / ConversationRelay), TTS/STT, sub-200ms latency pipelines
• Vector DBs — Qdrant, pgvector, ChromaDB, Pinecone
• Backend — Python, FastAPI, .NET / C#, Node.js / TypeScript, REST APIs
• Data — NLP-to-SQL (Vanna.ai), PostgreSQL, ETL, Data Pipelines
• Frontend — Vue 3, React, Dashboard / Data Visualization
• DevOps — Docker, GitHub Actions, AWS (Cognito, S3, ECR), Git

If you need a Claude agent, a voice AI pipeline, a RAG backend, or a .NET system
modernized and AI-enabled — I've shipped all of these and documented the work.
Happy to answer any questions you may have.

Steps for completing your project

After purchasing the project, send requirements so Oguz can start the project.

Delivery time starts when Oguz receives requirements from you.

Oguz works on your project following the steps below.

Revisions may occur after the delivery date.

Analyze the Legacy System

Review the provided legacy system, APIs, services, or database schema to understand business logic and user pain points.

Define MCP Contracts and Contexts

Design Model Context Protocol (MCP) contracts to wrap and abstract the legacy system’s capabilities.

Review the work, release payment, and leave feedback to Oguz.