Case Studies

Case Studies: AI Projects Running in Production

Not slideware — systems running in production. Each solution is also available as a ready-made product we adapt to your company. We keep client names confidential.

B2B · Multi-industry

AI agents that find new customers for sales teams

For companies across several industries, we designed and deployed AI agents that identify and qualify prospects and hand them to the sales team as ready-to-work opportunities.

70%increase in customer acquisition rate

Available as a product

The challenge

Prospect research was done manually, consuming a significant share of the sales team’s time.
Prospects were not assessed against shared criteria, so time was spent on poor-fit companies.
First-contact messages were generic and follow-ups were not carried out consistently.

What we built

A targeting framework that defines the ideal customer profile by industry, size and need signals
Research agent: identifies companies matching the target profile, and the right contacts, from publicly available sources
Qualification agent: scores and prioritizes every prospect against shared criteria and hands it to sales with the rationale
Outreach agent: drafts personalized first-contact and follow-up messages for each prospect and manages the follow-up schedule
A control layer that sends every message for human approval, processes opt-out requests automatically and logs the process

Results

Customer acquisition rate increased by about 70%.
The sales team now focuses on qualified prospects instead of research.
The same setup was quickly adapted for companies in different industries.

Industry · Manufacturing

Automating public tender (EKAP) processes with AI

We automated tender tracking and process management on EKAP, Türkiye’s public procurement platform, for manufacturers bidding on public tenders.

90%of manual work automated

Available as a product

The challenge

Tracking tender notices on EKAP, picking relevant tenders and recording them was done by hand.
Submission dates, bids and outcomes were processed manually and took several people’s time.
There was a constant risk of missing critical deadlines or document requirements.

What we built

A tracking module that regularly scans tender notices on EKAP and filters those matching the company’s line of business
An AI assistant that reads administrative and technical specifications and extracts a summary, required documents and key dates
Automatic reminders for submission, tender and award dates
A dashboard that brings participation, bid and outcome data together in one place

Results

About 90% of the manual steps in the tender process are now automated.
A single person can now manage the whole process with ease.

Industry · Manufacturing

Digital workforce: AI agents that run operational processes

For a manufacturing company, we designed and deployed a multi-agent AI system that continuously monitors operational processes and takes on defined responsibilities as a member of the team would. The architecture adapts to any industry that relies on process tracking.

Digital workforce

Available as a product

The challenge

Order, work order and task status was tracked manually across different departments and scattered sources.
Process deviations and delays were detected late, leaving little room for timely intervention.
Status reports were compiled by hand, limiting management’s real-time operational visibility.

What we built

A multi-agent architecture in which coordinated AI agents each hold a defined role and scope of authority
Process monitoring agent: continuously tracks open orders and work orders and identifies deviation and delay risks early
Coordination agent: escalates items that need action to the responsible owner and follows up on upcoming and overdue work
Reporting agent: automatically prepares daily and weekly operations reports for management
Governance layer: routes critical decisions for human approval and keeps an auditable log of every agent action

Results

Routine tracking, notification and reporting are now carried out by the agents.
Deviations and delays are escalated to the responsible owner as soon as they occur.
Management has up-to-date operational visibility without waiting for reports.

E-commerce · Food

AI agents answering requests via WhatsApp, email and a website chatbot

We automated customer communication for a food e-commerce company with AI agents.

80%lower team workload

Available as a product

The challenge

Customer messages arriving via WhatsApp and email were answered one by one, by hand.
During busy periods replies were delayed and the team was swamped with routine questions.
Requests arriving outside business hours waited until the next day or were missed.

What we built

An integration that brings WhatsApp, email and the website chatbot into a single flow
A multi-agent architecture that classifies each message by topic and priority and routes it to the right agent
An AI chatbot on the website that answers visitors’ questions instantly and logs their requests
AI assistants that instantly answer common questions using the company’s own information
A human-in-the-loop step that hands complex or sensitive requests to the team with full context

Results

The team’s customer-communication workload dropped by about 80%.
Customer requests and incoming messages are answered instantly, including outside business hours.

AI Assessment

We assess your processes and data landscape together to identify where AI will create the most value for your organization, and define a clear implementation roadmap and data-security requirements.

Request an assessment WhatsApp