AI State of the Art

Autonomous agents
hacking organizational capabilities

Beyond the hype

Is there an
AI bubble?

The debate around a potential AI bubble is intensifying. While there is no definitive answer, several signs suggest the possibility: the tech sector is seeing enormous investments in AI, with startups reaching high valuations rapidly. However, some experts believe progress with current language models is slowing, and expectations for Artificial General Intelligence (AGI) are being revised, with industry leaders postponing their forecasts and lowering the bar.

Despite uncertainties, it’s clear that AI has already transformed
the world. Even if the bubble bursts or technological advances
stall, the changes are irreversible.

Everyday tasks, such as document creation, research, and automation, have been revolutionized, fundamentally changing workplaces and society.

The real challenge is not predicting the fate of the AI bubble but adapting to the new reality AI has created. Organizations and individuals should focus on leveraging these advancements and evolving their work practices, rather than waiting to see if the current momentum will continue or collapse.

AI is advancing rapidly, with innovation concentrated in early-stage technologies that promise significant transformation. The defining goal for organizations is not the proliferation of solutions but ensuring robustness, reliability, and measurable impact.

Strategic success in AI requires a shift from trend-driven
adoption toward building strong foundations.

This involves prioritizing technologies that enable scalable, efficient, and secure deployment, supported by disciplined risk management, robust data management, and realistic expectations. Expanding data management capabilities is critical to ensure high-quality, well-governed data that can sustain advanced AI models and prevent issues related to bias, reliability, and compliance.

Technologies

Secure deployment

Scalable

Efficient

AI strategy

STRATEGIC ALIGNMENT

The differential
resonance of AI

The progression of AI creates challenges that resonate differently across industries. For strategic success, organizations must identify prevalent concerns within their sector, recognize practical examples, and leverage these insights to derive value from AI.

By systematically analyzing cognitive frictions, tedious or repetitive tasks that limit organizational processes, organizations can justify AI usage only where it compensates for genuine bottlenecks.

AI success:
the three outcomes that matter

Business

Achieving the right benefits, such as increased productivity and revenue generation, at the right costs.

Technology

Securely managing and securing the data, models, and applications that underpin the AI systems.

Behavior

Addressing how employees feel about AI and how leaders must evolve their management approach.

High
suitability

Medium
suitability

Low
suitability

autonomous Intelligence

The need for
composite AI

Although generative models are highly suitable for content generation and conversational user interfaces, they are not ideally suited for essential business functions such as prediction/ forecasting and planning.

These core functions rely heavily on non-genetic machine learning, optimization, simulation, and graphs. This is why organizations should transition toward Autonomous Intelligence to guarantee the robustness and reliability of their systems.

This strategy involves deploying systems that combine GenAI for interfaces and content creation with nongenerative techniques for core analytical processes.

Real applications
for our clients

Here we will present some of the solutions developed for our clients. They are grouped according to the use-case families they address and the specific AI techniques most suitable for these families.

Real Cases

Conversation
user interfaces

Generative AI Machine Learning Graphs Heuristics

Agent for education

On-demand support for students and teachers

Grapevine

Conversacional agent to chat with a database.

Agent for fintech

Report automation via multi-agent platform

Quikmed

AI-enhanced patient care

Check some of the conversational solutions we built for our clients:

Anomaly detection
& monitoring

Machine Learning Graphs

Check the monitoring solution we built:

Intelligent
automation

Machine Learning Graphs Generative AI

Ferracioli

Accounting orchestration and automation:
from 4 analysts to 1 supervisor with GenAI

Content
generation

Generative models Simulation

Perception

Generative models Machine Learning

Segmentation

Machine Learning Rules/heuristics Graphs

Wunder

Agent to automate the initial management
and pre-qualification of automotive credit
applications through WhatsApp.

AI techniques heat map

The heat map indicates the intrinsic strengths of different
AI technique families across common use-case functions.
Source: Gartner 2025

USE-CASE FAMILIES	COMMON AI TECHNIQUES
USE-CASE FAMILIES	GENERATIVE MODELS	NONGENERATIVE MACHINE LEARNING	OPTIMIZATION	SIMULATION	RULES/ HEURISTICS	GRAPHS
Prediction / forecasting	L	H	L	H	M	L
Planning	L	L	H	M	M	H
Decision Intelligence	L	M	H	H	H	M
Autonomous systems	M	M	H	M	M	L
Segmentation	M	H	L	L	H	H
Recommendation systems	M	H	M	L	M	H
Intelligent automation	M	H	L	L	H	M
Anomaly detection/ monitoring	M	H	L	M	M	H
Content generation	H	L	L	L	L	L
Conversational user interfaces	H	L	L	L	M	L
Knowledge discovery	H	M	L	L	M	H
Perception	H	H	L	L	L	L

H High suitability

M Medium suitability

L Low suitability

Free access to the 20-page PDF

This report features

AI governance: The 4 core principles for enterprise-level deployment and risk management.
The path to autonomous intelligence: A framework for transitioning from siloed data to proactive, agent-driven insights.
Model distillation: Strategies to reduce latency and costs by training “student” models from “teacher” models.