Marketing automation through AI

Budget: 7000 EUR Deadline: 43 days

Welcome! My name is Oleg, I am the project manager at Business Atlas. Creating complex multi-agent systems and autonomous digital ecosystems is our core expertise, so we clearly understand the difference between a linear bot and an engineering AI architecture. We are fluent in English for discussing tasks and synchronizations.
Here is our technical vision for implementing your specifications:
1. Multi-Agent Systems and RAG
• We build autonomous workflows in n8n/Make with the logic of interaction among several agents (Multi-Agent Systems), which pass data along the chain, log steps, and perform cross-validation (one generates, the other verifies).
• Our tech lead Lavr has deep expertise in integrating LLM and building RAG systems so that AI clearly distinguishes between new context and knowledge base.
2. Marketing Automation and Deep Research
• Data collection and deduplication: We will set up data structuring, duplicate removal, and reconciliation with the article database. To bypass API limitations of social networks (X, Instagram, LinkedIn) and analyze metrics, we will integrate browser parsing (Puppeteer/Scraping API).
• Content factory: Our solutions can analyze the brand's Tone of Voice and create tailored posts for platforms with AI moderation before the final report.
3. Interface and Analytics
• For transparent control, we develop no-code interfaces or dashboards where you can see forecasts, token costs, detailed logs of agent operations with prompts, as well as edit and approve generated posts.
We suggest starting with an expert diagnosis to detail the semantic and metric criteria for analysis and design the architecture with ROI calculations. Message me privately to discuss the details! Expand

Budget: 350 EUR Deadline: 3 days

I specialize in exactly this — I already run such a system in production:
research → dedup-against-history → multi-agent self-check → styled posts.
So rather than build from zero, I'd adapt my existing framework to your
case (your sources, style, platforms) — faster and proven.

For 350 EUR I can deliver a working Phase 1: sources + dedup-against-history
+ context analysis + report + one platform's posts. Anything beyond that
scope — hourly (50 EUR/h).

I work in writing (English) — for a spec this detailed it keeps every
requirement documented and nothing lost between us.

Two questions:
1. Your deep-research sources — sites, RSS, or APIs? Which social platforms?
2. Scraping social for likes/comments needs proxies/anti-detect to avoid
account bans — whose accounts and proxies do we use?

If text-based collaboration and a scoped Phase 1 work for you, I'm ready to start. Expand

Budget: 5000 EUR Deadline: 5 days

Good day! I build exactly such multi-agent systems — not a one-time bot, but a network of agents, each with its own role, passing data to each other and checking the neighbor's result before outputting.

According to your flow, the logic is clear: deep research on specified sources, deduplication of new data against what has been found earlier (through RAG/vector memory, to avoid pulling duplicates), cross-checking with already published articles, assessing the importance of topics in the required context, a report for you, and drafts for LinkedIn / Instagram / X in your style with final self-checking.

Stack: Python, orchestration of several agents, LLM calls, MCP tools, RAG for memory and deduplication. I already have a working multi-agent system that conducts research daily, removes duplicates, and prepares content for social media — I can show a live demo so you can see the real flow, not just promises.

I suggest starting with a small pilot: one working chain of agents (research sources -> dedup against memory -> draft post in your style with self-checking), so you can assess the quality in practice, then we can scale up to full connectivity. The starting price is attractive — I'm gathering initial feedback here, so I'm offering a good deal. Approximately 4-5 days for the pilot.

One clarification to hit the mark: are the sources for research (sources 1,2,3...) specific websites/RSS, APIs, or rather topics for search? And where is the archive of already published articles located, against which to check for duplicates?

Ready to start immediately. Expand

Budget: 2000 EUR Deadline: 7 days

Hello, I have worked on AI agents for automating content marketing with a RAG system and deep research functionality - the project increased content creation efficiency by 340% in 2 months.

I am curious about which LLM model you plan to use as the main one for your agents and whether you are considering multimodal solutions?

I suggest we get in touch; I will provide you with free technical consultation and we can outline a development plan + I will tell you about my team!

Budget: 200 EUR Deadline: 7 days

Good day. I work on this every day: I have my own multi-agent AI system in production — 100,000+ lines of Python (FastAPI), 750+ API endpoints, 1,000+ automated tests.
What is implemented in it (all — custom code, I will show it live):
Orchestrator: 19 specialized agents, tasks as a dependency graph — parallel waves.
Agents: Planner → Specialist → Reviewer → Critic → Finalizer.
Model router: model selection based on task complexity, multi-provider support (OpenAI-compatible APIs, Gemini, Groq, DeepSeek) with automatic fallback, accounting for quotas and model success rates. Integrated Telegram Discord bots.
I am ready to show the system live.

Budget: 500 EUR Deadline: 15 days

I will create the system in separate parts: one searches for new topics from your sources, the second checks them for duplicates (comparing with what has already been found and what has already been published), the third evaluates the topics based on criteria we will agree on, and the best ones proceed for further processing, while the last one prepares posts for LinkedIn, Instagram, and X in your style, taking examples from previous posts.

Where the official API is too limited, I will take likes and comments directly from the page through the browser, just as I currently collect similar data where official APIs do not provide what is needed. Duplicates are filtered based on the database of what has already been processed, so nothing is repeated, and every step, including prompts, models, and costs, is logged for the dashboard you described.

Stack: LangGraph for agents, vector database for RAG based on old articles and topics, Playwright for data collection from social networks, simple web interface for logs, forecasts, and ready posts. Examples of work with agents and RAG are in my profile.

Details on the list of sources and criteria will be discussed in the chat.

Budget: 400 EUR Deadline: 10 days

Hello, Mikhail!

You described exactly what I do: an AI agent as a pipeline — input processing → LLM call → tools via MCP if necessary → RAG on the knowledge base → output processing and validation (not just "prompt in chat").

In production, I have such an agent: the orchestrator routes the request, accesses tools (searching for products/data, actions), responds strictly from the knowledge base with an honest "I don't know" instead of hallucinations, plus logging and quality control of responses. So the entire cycle input→LLM→tools/MCP→RAG→output is not theory, but a working system.

Regarding "marketing automation through AI," I see options: an agent for generating/personalizing content from your database, qualifying and responding to leads, assembling reports/insights from data. What is the priority — content, leads, or analytics?

Please describe the task in more detail (data sources, what actions the agent should perform, where to run it — in the cloud or locally) — I will propose a specific architecture, timelines, and costs. I am ready to discuss the details in private.

Budget: 4000 EUR Deadline: 15 days

Hello Michael,

Your project is very close to what we build at RAI (Robots Artificial Intelligence).

We develop AI agents, multi-agent workflows, RAG systems, knowledge bases and business automation solutions. One of our core products is based on orchestrating multiple AI agents that collect information, validate results, access knowledge bases and generate structured outputs.

From your description I see several important blocks:

• Deep Research from multiple sources
• Deduplication and relevance scoring
• Multi-agent validation workflow
• RAG-based contextual analysis
• Content generation for LinkedIn, X and Instagram
• Logging, traceability and cost tracking
• Dashboard for reviewing outputs and prompts

I would recommend implementing this as a modular architecture where separate agents are responsible for:

1. Research
2. Data validation
3. Deduplication
4. Context enrichment through RAG
5. Content generation
6. Quality review
7. Reporting

This approach scales much better than one large agent and allows replacing models or data sources without rebuilding the entire system.

English communication is not a problem.

I would be happy to discuss the data sources, expected volumes, update frequency and preferred architecture.

Best regards,
Tadeush Muratov
CEO, RAI (Robots Artificial Intelligence) Expand

Budget: 3650 EUR Deadline: 33 days

Good day, Mikhail.

English: B2

Architecture: the agent-researcher gathers topics from your sources through deep research, the RAG layer on pgvector stores the findings and publications for deduplication through semantic search, the agent-analyst assesses relevance and importance in the context of your niche, the agent-copywriter generates posts for LinkedIn/Instagram/X in your style, the agent-reviewer checks for style compliance and factual accuracy.

Agent orchestration through LLM cycles with mutual verification: each result is validated by the next agent before passing it further. The style of posts is learned from your previous publications using few-shot examples from the vector database. The final report consists of ready posts with an evaluation of topics that you approve or adjust.

Architectural safeguards (Quality and Safety):
- Protection against hallucinations: Self-RAG pattern with cross-validation of facts. The agent-reviewer automatically returns the post for regeneration if the final text contains facts not present in the sources.
- Combating Context Poisoning and Subject Drift: Strict isolation of context through State Machine (transmitting strictly structured JSON instead of the entire dialogue log). Validation of input data for Prompt Injection before writing to RAG.
- Token optimization: Hybrid architecture — primary scoring and deduplication on fast/cheap models, final copywriting and review — on top-tier models (Sonnet/GPT-4o).

Stack: Go (provides parallel processing of sources through goroutines and low system overhead) for orchestration, PostgreSQL + pgvector for RAG, LLM providers Anthropic/OpenAI/OpenRouter, etc. MCP for integration with external tools. Deployment in Docker.

PM: Source files and documentation included. Expand

Budget: 2600 EUR Deadline: 16 days

👋 Good afternoon, the best and largest projects —> Freelancehunt

I have studied the task — it is a full-fledged multi-agent AI system with an orchestration layer, RAG architecture, and a validation contour for results between agents.

I have experience in developing similar AI pipeline systems, where chains are built: data gathering → filtering → semantic analysis → RAG enrichment → generation → self-check → structured output.

🔥 Here’s how I propose to implement it:

• orchestration layer for managing multiple AI agents
• search agent (web + social networks + deep research sources)
• RAG system considering history and previously published content
• deduplication module + semantic scoring of relevance
• analytical agent with logging of all steps and decisions
• content generation for LinkedIn / Instagram / X
• cross-validation between agents for quality control

The architecture is modular, extensible, with the ability to add new agents without changing the core of the system. I also plan to include observability (logs, traces, prompts, model outputs) to ensure the entire chain is transparent.

⏳ Deadline: 12–16 days
💶 Budget: 2200–2600 €

🤝 I am ready to discuss the architecture in more detail and propose the optimal option for your data sources and load. Expand

Budget: 1000 EUR Deadline: 1 day

Hello! The task with the multi-agent system is close to me — I really enjoy building pipelines where agents process input data, call LLM, go to RAG and MCP, and then check with each other and pass the results further. The scenario with deep research, deduplication of topics, and generating posts in style for LinkedIn, Instagram, and X is logically structured. I would be happy to discuss the details in private messages.

Budget: 5000 EUR Deadline: 30 days

Hello!

I am a Full-Stack Software Engineer with over 7 years of experience in developing websites, SaaS solutions, complex web platforms, and MVPs for startups - from idea and architecture to production and support.

I work not only as a developer but also with a focus on business logic, scalability, and long-term support of solutions. My portfolio includes examples of implemented projects of varying complexity.

Technology stack:
PHP (Laravel, Symfony, Yii2),
Frontend: JavaScript (Vue.js, React.js), HTML5, CSS3,
Databases: MySQL, PostgreSQL.

Regarding timelines and costs - at this stage, it is difficult to accurately assess them without clarifying the requirements. To form a reasoned estimate, it is usually necessary to understand:
- project goals;
- key functionality;
- expected load and integrations;
- priorities (quick launch vs scaling).

I suggest starting with a brief clarification of the requirements or a call/email - after that, I will be able to provide a realistic estimate on timelines, budget, and implementation options.

I would be happy to discuss the details. Expand

Budget: 100 EUR Deadline: 7 days

Hello, Mikhail! A person, not a bot :) I am building multi-agent systems in production: input data → LLM → MCP/tools → RAG → cross-checking by agents → report. A live example is an AI consultant for an online store (PydanticAI + OpenRouter, RAG by catalog, competitor monitoring through embeddings + LLM-grader). Your case with deep research and posts in style is exactly this architecture. Conversational English is fine. The price in the bid is conditional — negotiable, ready for a call.

Budget: 7400 EUR Deadline: 15 days

Hello,
I hope you are doing well.
I am very interested in your project involving marketing automation through AI, and I am ready to take on the job.
I have a strong background in data analysis, technical workflows, and structuring complex information. I closely follow AI technologies and understand how to leverage modern tools to optimize and automate data-driven processes efficiently.
I am ready to start immediately. Please send me the details and specific requirements so we can discuss the next steps.
Best regards,
Yurii

Budget: 25000 EUR Deadline: 14 days

We already have a nearly ready solution with AI agents and RAG, which can be quickly adapted for marketing and discussed here on the marketplace - I'm available (:
For the first stage, I would allocate 96,000 UAH and 14 days - this is a prototype of the agent chain, where the input data goes through a language model, RAG, MCP if necessary, self-check, and a report for the user.
We can keep it simple at the start - first, we gather 2-3 sources, a database of published materials, brand tone rules, and duplicate checks, then we add LinkedIn, Instagram, and X.
Am I correct in understanding that you already have an archive of articles and examples of posts to learn the style from?
The second question is - which sources for deep search should be considered primary and how often should the agent be launched?
A similar case on AI and content generation - https://business.ingello.com/vorfahr
A similar case on agent automation development - https://business.ingello.com/fractal
The main page of Ingello for FLH - https://systems-fl.ingello.com
If we proceed carefully, the result of the first stage will be a working prototype, a data schema, evaluation rules for topics, and a list of risks for industrial launch.

Budget: 13700 EUR Deadline: 10 days

Hello! Yes, the task is clear. We can take on such a system.

We have experience in AI automation, building workflows with LLM, working with RAG logic, generating marketing content, and creating systems where the result goes through several stages of verification before being delivered to the user.

I should note right away: this is not a simple task at the level of "connecting a bot." A proper architecture is needed: sources, search, deduplication, analysis, history storage, verification with already published materials, report generation, posts, and an interface for process control.

We can discuss the task in English. I suggest we go step by step: first, an MVP with a limited number of sources and social networks, then expand the system.

MVP estimate: 6–8 weeks, budget from $8,000–15,000. The full version with an extended interface, logs, RAG, browser automation, analytics, and multi-agent logic will be evaluated separately after discussion.

Budget: 350 EUR Deadline: 7 days

Hello. I understood the task: I need a person who actually builds AI agents, not just tweaks a single model. According to the description, this is a multi-agent system: processing input data, queries to LLM, calls through MCP, RAG for context, and several agents that pass data to each other and verify one another. Plus, an applied scenario for marketing: deep research on sources and social media, cross-checking with what has already been found and deduplication, analysis of topic importance based on criteria, a report with graphs, and generation of posts for LinkedIn, Instagram, X in a specified style with verification, and an interface with logs, requests, prompts, and costs.

This is exactly the area in which I work every day. I lead agency development through the orchestration of several agents with verification cycles, I have deployed RAG in production (semantic search via Qdrant with embedding models), and connected MCP tools so that the agent works with real sources and data schemas, rather than making things up. Multi-agent data transfer and mutual verification is a working pattern for me, not a theory.

For your scenario, I see it like this: a task orchestrator, a research agent with deep research on sources and social media searches through a browser where there is no API (considering likes and comments as metrics of importance), a RAG and memory layer to store what has already been found and published and to remove duplicates, an analytics agent to assess importance based on semantics and metrics, and a post generator for each platform with its own style plus a separate verification pass. Everything with logging of requests, prompts, and costs in the interface.

To aim accurately: what sources need to be monitored and which social media are a priority, and what models do you plan to work with, is there a preference for the provider?

Budget: 345 EUR Deadline: 1 day

Automatic response to the incoming application based on RAG systems criteria received. Please contact urgently within 24h.

Budget: 18000 EUR Deadline: 7 days

Hello. The AI part here should be done not as "ChatGPT magic," but as an engineering system with rules, logs, and verifiable results.

What I understood about the task: I am looking for a person (not a bot) who understands AI agents and knows how to build them. Also, an understanding of RAG systems is needed.

Here’s how I would approach it:
1. First, I would limit the AI with understandable input data, rules, and a verifiable response format.
2. I would add a deterministic layer: validation, business rules, logs, fallbacks — so the system does not rely solely on the model.
3. Then I would connect OpenAI/LLM as an analytical layer that explains and helps make decisions.

Relevant experience: Recently, I built a production system with n8n + PostgreSQL + OpenAI + Telegram: uploading Excel/Drive, validation, daily tasks, summaries for the manager, and retry logic.
Regarding timelines: the estimate is 7 days. For the budget: 18,000 — after a brief clarification of the technical specifications, I can confirm it more precisely.

If relevant — in the first message, I will provide a short implementation plan and a list of clarifications to immediately fix the MVP without unnecessary development. Expand

Budget: 350 EUR Deadline: 10 days

Hello. I read carefully — this is exactly what I have been doing for the last six months.

I built a multi-agent system for a law firm: RAG based on Ukrainian legislation, a validation agent that checks the responses of the second agent before delivering them to the client, and context transfer between steps through structured output. Quality metrics through RAGAS — a baseline is already in place.

Regarding your case with marketing automation: I understand the task specifically. Deep research → deduplication against history → contextual analysis → post generation in the required style + verification — this is a classic pipeline with memory and a self-check agent. I did something similar for a content strategy, where the agent compares new content with already published content and rejects duplicates.

Stack: Claude API (Anthropic), MCP servers, vector databases for RAG, Python/FastAPI. I do not use no-code platforms — only code, which means full control over the agents' logic.

What specific sources do you have for deep research? (websites, RSS, API?) — this will determine the architecture of the first agent.

Budget: 25 EUR Deadline: 1 day

Good day, Mikhail! In general, the task is clear, but to provide an accurate response regarding the deadlines and price, I would like to clarify some questions that arose after analyzing your task. Please write in private messages — we will discuss the details and your wishes.

Budget: 12345 EUR Deadline: 3 days

Hello. Are you already using any ready-made tools for agent orchestration, or do you plan to build everything from scratch?

I will clarify the details regarding timelines and budget in personal correspondence.

Here’s how I will execute this project:
1. I will design an architecture with several agents, where each performs its function: deep research, deduplication, context analysis, and content generation.
2. I will implement a data transfer chain between agents with automatic result verification and duplicate removal at each stage.
3. I will integrate a RAG system for storing search history and already published articles, as well as MCP queries for external sources.

Thank you for considering my proposal. I look forward to the opportunity to collaborate with you!

Budget: 300 EUR Deadline: 3 days

Hello. I created AI agents that worked with emails and social networks. They performed various tasks such as searching, analyzing responses to questions, creating posts, and more.

If you're interested, feel free to write, and we can discuss.