Erich – AI agent development, Python, OpenAI

An AI voice agent platform for healthcare sales and customer service in the U.S. market, enabling businesses to run conversational AI agents that handle live customer calls at a scale of more than 15,000 calls per day. The platform supports lead qualification, customer inquiries, and service or sales interactions through LLMs, speech-to-text, and text-to-speech technologies.

• Built and maintained AI voice agents for healthcare sales and customer service operations in the U.S. market;
• Developed conversational AI systems handling more than 15,000 customer calls per day using LLMs and STT/TTS technologies;
• Connected AI agents with internal business systems to automate lead qualification, CRM updates, and customer workflows;
• Integrated AI agents with MCP-based tools to retrieve information and perform real-time actions during conversations;
• Designed prompts, conversation flows, and evaluation frameworks to improve reliability, customer experience, and business outcomes.

An AI-powered e-commerce platform that turns a single product photo into a ready-to-sell product catalog, including a professional catalog-style image, title, description, and structured product details. It helps merchants automate a manual catalog creation process, making product preparation faster and more scalable. The platform reached 200 users within the first two months of launch.

• Developed multimodal computer vision workflows that analyzed product images and automatically generated titles, descriptions, and structured product details;
• Built a generative image system that transformed raw product photos into professional catalog-style images using reusable style templates;
• Designed prompt templates and multimodal guardrails to ensure generated descriptions remained accurate and grounded in the product image;
• Grew the platform to 200 users within the first two months after launch.

An AI-powered outbound calling platform for telecom sales and debt collection, enabling businesses to automate customer conversations through real-time voice agents. The platform handled more than 5,000 customer calls per day, achieving conversion rates comparable to or higher than human agents while significantly reducing operational costs.

• Designed and deployed AI voice agents for telecom sales and debt collection campaigns, handling over 5,000 real customer calls per day;
• Achieved conversion rates comparable to or higher than human agents in several campaigns while significantly reducing operational costs;
• Built real-time conversational systems integrating LLM dialogue, speech recognition, and voice synthesis to automate outbound calling workflows;
• Designed prompt templates, behavioral guardrails, and conversational frameworks to ensure natural and reliable AI interactions;
• Developed a UI-based configuration system allowing non-technical users to create and customize AI agents without writing prompts.

A sales intelligence and prospect research platform that helps businesses identify target companies and the right decision-makers, with verified contact details, for outbound sales and marketing campaigns across the U.S. and Europe. The platform combines large-scale company, hiring, and technology data into structured prospect intelligence and analytics used by Sales and Marketing teams for campaign planning and targeting.

• Led the company’s Research Department, building data-driven systems for outbound lead generation;
• Developed large-scale web scraping pipelines to collect company, hiring, and technology data from sources such as LinkedIn, Glassdoor, and SEC filings;
• Designed and deployed automated data pipelines that generated a dataset of 200,000+ companies and their decision-maker contacts, including verified emails and phone numbers;
• Delivered prospect intelligence and analytics dashboards used by Sales and Marketing teams to target outbound campaigns across the U.S. and Europe.

A data-driven platform for vehicle sales and servicing that helps teams improve customer data quality, understand customer behavior, and support proactive service engagement. It combines data enrichment, analytics, customer feedback, and predictive insights to help commercial and operational teams make better use of customer information.

• Generated databases for the company's migration to the SAP ecosystem by implementing ETLs for data extraction, wrangling, cleaning, and formatting;
• Developed Python scripts for counting client visits and in-store tracking using security camera feeds and computer vision;
• Designed, prototyped, and validated an in-store survey to measure NPS for visiting clients using tablets and custom dynamic surveys;
• Collaborated with and supervised designers and software providers for national deployment of the devices;
• Developed Tableau dashboards for the data collected from the surveys;
• Improved data accessibility by designing and building a better data structure alternative to the SAP database;
• Extracted, cleaned, and transformed SAP data with SSIS and Python;
• Proposed and created new ways to generate leads by applying web scraping to social networks, competitor websites, and other internet sources;
• Used generated leads for mailing and calling campaigns.