Hubert Zygmunt

Processes that drain your people - I hand them over to AI.

I focus on what produces measurable results. Not on single tasks, but on entire systems and processes.

See what I do

What I do

I help B2B companies free up their team's time, increase efficiency, and build brand visibility online. Through two complementary tracks.

Automations and AI agents

I build solutions that take over repetitive processes across different areas: customer service, sales, marketing, and operations.

Dla kogo

For small and mid-size companies, regardless of industry. AI tools don't care about industry, and problem categories repeat, so what works in manufacturing also works in services.

Example areas

  • RFQ qualification and quoting support
  • AI assistants and chatbots based on your knowledge base
  • Customer service automation
  • Sales process support
  • Marketing and operations automation
  • System integrations and data flow automation

* Every solution is built individually for the client's problem. The above are examples - I create them based on company needs.

Ask for a quote

B2B content marketing

I run LinkedIn profiles for CEOs, decision-makers and specialists at B2B companies. Their activity translates into brand visibility among the target audience. I specialize in technical industries where specialist content builds trust faster than ads.

Dla kogo

For small and mid-size B2B companies in technical industries (manufacturing, industry, automation) that want to build visibility and personal brands of decision-makers among the target audience - translating into measurable business effects.

Example areas

  • Current state audit and action plan
  • Running LinkedIn profile
  • Creating longer specialist content (articles, publications)
  • Coherent visual style of publications
  • Building company knowledge base
  • Monthly and quarterly reports with performance analysis

* All these areas are tailored to the company's specifics, target audience and business goals.

Ask for a quote

How I work

Every project, whether AI implementation or content marketing, goes through the same proven process.

01

Consultation

Free 30-minute conversation where we talk about your company, problems, bottlenecks, and expectations. The goal is to check whether I can help solve your problem. If I can't help, I say so directly.

02

Audit and proposal

I conduct a full audit of the current state, examining what systems and tools you use, what processes operate in the company, and analyzing the target audience or buyer personas if we're talking about content marketing. Then I propose a specific solution, and you decide whether we move forward.

03

Implementation

For technology solutions: I set up the workflow, integrate with your systems, prepare the knowledge base. For content marketing: I set publication rules, target audience, prepare company knowledge base, visual style proposal, and first materials.

04

Testing and calibration

For technology solutions: agent testing on real company data, fixing reported bugs, iteration until it works as it should. For content marketing: first publications, hour and format tests, calibration based on first results.

05

Support and growth

After deployment I stay on board. Collaboration can be ongoing or project-based, depending on needs. For technology solutions: ongoing support, bug fixes, expansion into new areas. For content marketing: continued publications, content adjustments based on results, topic diversification, strategy growth.

Projects

B2B content marketing on LinkedIn - PRASMET
+166% Post reach growth

Problem

PRASMET is a Polish manufacturer of serial metal components with over 20 years of experience and a reputation among clients in the energy, automotive, lighting, and home appliance sectors. Despite a strong technical position and collaboration with major global corporations, the company's presence on LinkedIn was sporadic - publications appeared occasionally and content had a "company catalog" character. Reach was low, even though decision-makers at end clients increasingly look for suppliers on LinkedIn before sending an RFQ.

Solution

We built a content production process that unloaded the client - I handle everything, the CEO is responsible only for substantive verification. We work based on the company's knowledge base, clear content creation rules, and a publication schedule tailored to business priorities. All results are analyzed in monthly and quarterly reports, so the content is constantly being refined.

Effects

+166% average post reach growth
+233% average comments per post growth
+324% total reach growth
~85k record views on a single publication
  • Stable rhythm of three regular publications per week, independent of the client's calendar and availability
  • CEO unblocked - focuses solely on substantive verification, no longer inventing, writing or planning content
  • Style change from company catalog to content from the client's perspective - topics that actually interest them
  • Direct RFQ generated from a single publication - confirmation of pre-sales value
AI agent for RFQ qualification - PRASMET
in progress Status

Problem

The company's inbox received dozens of RFQs per month, of which a large portion arrived with missing information needed for quoting. Manual analysis of each RFQ took several minutes, taking up time that could be used for talking to clients and actual offering.

Solution

I built an AI agent integrated with a dedicated Slack channel. The agent automatically analyzes each RFQ, researches the client company, and returns a structured summary for sales approval. Everything based on a company knowledge base (RAG) with relevant materials, which significantly reduces the risk of model hallucinations.

  • Agent integrated with a dedicated Slack channel
  • Company knowledge base (RAG) with relevant materials - reduces hallucinations
  • Automatic research of the client company from the RFQ
  • Additional features: CRM note, email draft to the client
  • * Planned development direction: lead scoring, email-based quote context aggregation, production reports, quality complaints agent

Current effects

<1 min analysis time per single RFQ
24/7 agent availability
  • Analysis time per RFQ reduced from minutes to under a minute
  • Every RFQ returns to the salesperson with ready context - client research, RFQ analysis, and proposed next actions
  • All bugs are fixed on an ongoing basis, taking into account client feedback during testing

Describe your problem

Free conversation up to 30 minutes during which you tell me about your company, the problems you face, and we discuss a potential solution. No sales agenda. If I can't help, I'll say so directly.