You will get audit and cut your AWS cloud costs
Project details
Your AWS bill is probably 30–50% bigger than it needs to be. I'll audit your account, find the waste — oversized EC2, idle resources, on-demand instances that should be Savings Plans, unoptimized S3 and RDS — and cut it without touching performance or uptime. You get a prioritized report with real dollar figures, and I can implement the fixes for you. I also optimize AI/ML workloads (Bedrock, SageMaker, GPU). Message me your monthly spend before ordering.
Project Type
Financial Consulting, Training & DevelopmentWhat's included
| Service Tiers |
Starter
$150
|
Standard
$600
|
Advanced
$1,500
|
|---|---|---|---|
| Delivery Time | 3 days | 8 days | 15 days |
Number of Revisions | 1 | 3 | 5 |
Frequently asked questions
1 review
(1)
(0)
(0)
(0)
(0)
This project doesn't have any reviews.
DM
Dmitri M.
Jan 20, 2026
Una Developer for System Configuration
The team delivered everything on time and on the agreed upon budget.
About Daniel
Agentic AI Developer | LLM Agents, RAG & Workflow Automation
Lviv, Ukraine - 6:02 am local time
Over the last few years my team and I shipped platforms like Revbox and Plusclix, powering millions of user interactions daily. I bring that same "make it scale, make it reliable" discipline to AI.
━━━━━━━━━━━━━━━━━
RECENT IMPACT (LAST 3 MONTHS)
━━━━━━━━━━━━━━━━━
AWS cost −88% — $5,000 → $600/mo via architecture refactor, right-sizing, and caching.
$25,000 in NVIDIA/AWS credits secured — 2+ years of infrastructure at current burn.
Shipped an open-source MCP SDK — onboarded 2 new publishers within weeks of release.
Productized AI — replaced manual taxonomy/tagline work with an LLM pipeline, sold as a $20/mo add-on.
Multi-agent dev workflow — Claude Code + Zed/Cursor, 2–3 agents in parallel, ~6× output vs. solo coding.
━━━━━━━━━━━━━━━━━
WHAT I BUILD
━━━━━━━━━━━━━━━━━
LLM Agents & Tool-Use — autonomous agents that call your tools, APIs, and databases. Function calling, MCP servers, multi-step reasoning, human-in-the-loop guardrails.
RAG & Knowledge Systems — assistants that answer accurately over YOUR data. Ingestion, chunking, embeddings, vector search, re-ranking, citations, and eval loops that keep hallucinations out.
AI Workflow Automation — LLMs wired into the tools you already use (Slack, Gmail, Notion, HubSpot, Airtable, internal APIs) via n8n, Make, and custom code.
Multi-Agent Systems — orchestrated agents that plan, delegate, and verify each other's work on long-running tasks.
Full-Stack AI Products — model integration to polished app: frontend, backend, infra, deployment.
━━━━━━━━━━━━━━━━━
STACK
━━━━━━━━━━━━━━━━━
Agents: LangGraph, LangChain, CrewAI, OpenAI Agents SDK, AutoGen/AG2
Models: OpenAI (GPT-4o/o-series), Anthropic Claude, Gemini, Llama, Mistral, fine-tuned
Protocols: MCP, function/tool calling, structured outputs
Vector DBs: Pinecone, Weaviate, Qdrant, Chroma, pgvector, Milvus
RAG: LlamaIndex, hybrid & semantic search, re-ranking, RAGAS evals
Backend: Python (FastAPI), Node.js/TypeScript, Postgres, Redis
Deploy: Docker, AWS, GCP, Vercel, serverless — with logging, monitoring, cost controls
━━━━━━━━━━━━━━━━━
WHY CLIENTS PICK ME
━━━━━━━━━━━━━━━━━
Production-grade, not proof-of-concept — systems that survive real traffic, edge cases, and cost limits.
A team behind me — you get engineering depth and continuity, no single point of failure, no "I'm on vacation" delays.
Business-first — I start from the outcome you need and work back to the architecture.
Clear communication — regular updates, honest timelines, documented handoff-ready code.
━━━━━━━━━━━━━━━━━
HOW I WORK
━━━━━━━━━━━━━━━━━
Discovery — map your use case, data, and success metric.
Prototype — a working slice, fast, so you can feel it before we scale it.
Build & harden — evals, guardrails, monitoring, cost optimization.
Ship & support — deployment, docs, optional ongoing maintenance.
Need AI built AROUND your workflow — not a generic ChatGPT wrapper? Message me with what you're trying to automate or build, and I'll tell you honestly whether an agent is the right tool and roughly what it takes.
Steps for completing your project
After purchasing the project, send requirements so Daniel can start the project.
Delivery time starts when Daniel receives requirements from you.
Daniel works on your project following the steps below.
Revisions may occur after the delivery date.
Connect read-only billing access
I send you exact IAM/Cost Explorer setup steps and confirm I can see your usage data.
Analyze account & usage patterns
I review spend across EC2, RDS, S3, data transfer and more to find every source of waste.