Innsbruck, Austria - December 20, 2024: In a world where operational complexity and fluctuating workloads challenge enterprises daily, forward-thinking organizations are embracing fully autonomous process automation. Last week, more than 80 invite-only participants from 15 different companies and partners tuned in to the 2024 End-of-Year Product Launch Event hosted by DeepOpinion—an agentic automation platform designed to help enterprises streamline complex back-office processes and high-volume knowledge work.
The event, themed “Zero to Autonomy,” offered a comprehensive look at how leading companies are leveraging native agentic automation to transform their operations and deliver tangible, customer-centric results. Attendees learned about new product features, gained insights into the evolving definition of “agentic” automation, discovered how to achieve unprecedented straight-through processing (STP) rates, and heard directly from customers who have already realized significant benefits. Most importantly, the event underscored that agentic automation is not just about technology; it’s about enabling people to refocus their efforts, streamline their processes, and confidently handle unpredictable demands—all while delivering a top-tier experience to their own clients.
Understanding Agentic Automation and the Path to “Zero to Autonomy”
DeepOpinion’s 2024 End-of-Year Product Launch set out to clarify the true essence of “agentic” automation by contrasting it with traditional RPA plus add-on AI approaches. While many solutions claim to be “agentic” by simply bolting AI-driven features onto existing robotic process automation frameworks, this approach often results in incremental gains, fragmented systems, and minimal straight-through processing (STP) rates—typically hovering between 0-10%. In contrast, native agentic automation is purpose-built from the ground up, designed to deliver end-to-end autonomy across complex knowledge work like claims management or purchase-to-pay cycles. This native foundation allows the platform’s integrated components to work in synergy, enabling STP rates of 80% or more and driving truly transformational results rather than mere task-level productivity improvements.
DeepOpinion, in their opening remarks, introduced the event’s theme: “Zero to Autonomy.” This concept envisions a journey towards a world of nearly self-governing operations.
For many companies, the journey starts at zero—where employees shoulder the burden of highly manual workflows—and moves toward complete autonomy, where the machine can handle the majority of tasks seamlessly and reliably. This journey isn’t just about plugging in advanced AI models. Instead, it’s about assembling the right combination of intuitive, no-code automation tools, domain-specific templates, performance metrics, confidence scoring, and human oversight at critical junctures. By bringing these elements together, organizations can redesign their processes to handle peaks and valleys, adapt to new scenarios, and free human workers from the monotony of repetitive tasks.
The True Business Value of Agentic Process Automation
Companies and partners attending the event were not simply looking for theoretical best practices; they wanted to see bottom-line impact. The Product Launch provided compelling business cases to show how agentic automation translates into measurable value.
A key driver is maximizing STP—straight-through processing—where an extremely high percentage of tasks are completed without human intervention. Attendees heard that when STP surpasses 80%, entire workflows can run near-autonomously. This shift not only reduces labor costs but also dramatically compresses timelines, accelerating delivery and ensuring that customers receive responses in near-real-time.
For example, one scenario presented at the event detailed a large-scale process handling 1 million manual cases per year. Traditionally, it took about 10 minutes per case, tying up roughly 100 full-time employees and costing around $3.8 million annually. By implementing an agentic automation approach and pushing STP to 80%, the same volume could be processed almost instantly—within 30 seconds per case—cutting costs down to $600,000 and saving $3.2 million. The difference is not just a number; it’s a tangible shift that allows enterprises to redeploy talent, improve service levels, and invest more heavily in strategic initiatives.
New Product Features: A Lifecycle Approach to Automation
To help organizations achieve these results, the event highlighted several new features available on DeepOpinion’s agentic automation platform. These enhancements reflect a mature understanding of the automation lifecycle—recognizing that real-world use cases rarely fit a neat template and often require iteration, oversight, and continuous improvement:
- Pre-Built Templates:
Enterprises no longer have to start from scratch. With over 100 pre-built templates tailored to common processes and verticals, teams can jumpstart their automation projects. By reducing time-to-value, organizations can realize efficiencies and ROI sooner. - Skill Versioning:
The introduction of skill versioning allows users to manage multiple iterations of their AI models. This approach makes it easy to test new instructions, compare performance metrics across versions, and deploy improvements confidently without risking production stability. - Performance Metrics & Ground Truth Testing:
Attendees learned that measuring accuracy and STP rates before going live is essential. By comparing predictions to known outcomes, organizations can validate their models and ensure they deliver the desired results right out of the gate. - Confidence Scores:
Trust is critical, and proprietary confidence scoring now goes beyond document-level checks. By assigning confidence levels at the individual data point, teams can skip reviewing the entire file and focus only on what matters. This level of granularity allows users to set specific thresholds for when human intervention is required, ensuring that attention is reserved for the most important elements. As a result, quality and compliance remain high—particularly in regulated sectors like insurance, finance, and healthcare—while reducing the time and effort spent on unnecessary manual reviews. - Human-in-the-Loop (Control Hub):
While the ultimate vision is full autonomy, human expertise remains invaluable for the entire end-to-end process. A single dashboard provides oversight not just for document reviews, but also for intermediate automated decisions, payment calculations, and even customer responses. This ensures quick human intervention when needed, without disrupting the overall automation pipeline. - Active Learning & Auto-Instruction Optimization:
The journey doesn’t end with deployment. Agentic automation platforms leverage active learning techniques to refine instructions based on real-world data. Over time, these systems become smarter, more accurate, and more attuned to the nuances of each enterprise’s processes.
Customer Success Stories: Real-World Wins
The highlight of the event was undoubtedly the chance to hear directly from customers who have transformed their operations using the self-service platform’s agentic automation capabilities. Two stories in particular were highlighted: CED Group and e&.
CED Group (Insurance Services):
CED Group deals with claims and expert assessments in the insurance domain, often facing massive spikes in workload due to climate events and other unpredictable factors. Before agentic automation, CED Group handled 50,000-60,000 monthly emails manually, leading to backlogs, slow response times, and employee frustration.
Emmanuelle Berthier, Group Chief Information Officer at CED, explained the challenge:
“The biggest challenge in this activity is managing fluctuating workloads. We handle claims and expert assessments, and when severe weather events occur—like storms—we see a surge in incoming tasks. Our operations aren’t stable; we have intense peaks followed by slow periods. This volatility makes it very complex to manage our operations effectively, since people don’t naturally work under such irregular conditions.”
By implementing agentic automation, CED was able to achieve an approximately 90% automation rate. As Berthier noted:
“Yes, we’ve managed to automate about 90% of our tasks. This means that now roughly 90% of our team can focus on more value-added, customer-facing responsibilities instead of basic, repetitive work.”
This shift not only improved operational efficiency but also elevated customer satisfaction through faster responses and consistent quality. Berthier cited three essential criteria for choosing their solution: seamless workflow integration, minimal infrastructure overhead, and robust reporting:
“In addition, I have three very important criteria. First, the solution must be fully integrated into our workflow, which is challenging given our very old legacy back-office system. Second, I want a solution that doesn’t need to be installed in our own infrastructure…I don’t want my team spending time maintaining it. With DeepOpinion, it’s completely managed, so we don’t need our IT staff handling ongoing issues. Finally, comprehensive reporting is essential. I need clear insights into what we’re doing, how many emails we receive, and so forth. We’re still fine-tuning these reports to ensure we have the data we need.”
Looking ahead, CED aims to go even further with automation, as Berthier shared:
“The next step is to take the process even further. For example, we want to automatically generate a dossier or a draft purchase invoice…having the draft ready would mean that most of the work is done, and human involvement would be limited to a quick review and final approval.”
e& (Telecommunications):
Another standout story came from e&, a world-leading telecom company. Their journey with agentic automation began with a critical focus on enterprise customer onboarding, a process previously hampered by complexity, manual intervention, and lengthy SLAs.
AbdulRahman AlMarzooqi, Sr. Manager of AI Cognitive Solutions at e&, summed up their objectives:
“As part of a technology company and department focused on AI and machine learning automation, our main objective is to continuously improve the services we offer our customers. We do this by adopting various cutting-edge technologies to enhance the customer experience, increase efficiency, and deliver real value.”
Time-to-market and accuracy were two foundational criteria for e& when selecting a solution:
“Our first key criterion was time-to-market. We needed a solution that could be implemented quickly…The second crucial criterion was accuracy. With so many customer-facing channels, any inaccuracies or misinformation directly affect the customer experience.”
Within weeks, they had an end-to-end automated onboarding solution running seamlessly across multiple channels—from mobile applications for business customers to web portals and internal operational tools. AlMarzooqi explained:
“With DeepOpinion, we achieved excellent results. It took only a few weeks to complete the entire end-to-end development and integration with our existing systems via APIs. We successfully implemented a fully automated onboarding solution across multiple channels.”
The impact was remarkable. Accuracy shot up to over 90%, while manual effort dropped by more than half. SLAs that once stood at 48 hours were now often measured in near real-time, eliminating delays and reducing human involvement:
“With DeepOpinion, we achieved remarkable results within just a few weeks…we improved accuracy to over 90%…The back-office team also became more productive, boosting efficiency by more than 50%. Moreover, we reduced our SLA from 48 hours to near real-time onboarding, eliminating the need for human intervention and delivering a seamless, touchless customer experience.”
Encouraged by these outcomes, e& plans to broaden the use of agentic automation into new business domains, from HR to legal documents:
“Based on the results we’ve seen so far, we’ve developed a roadmap to further expand…we’re currently planning to implement around 10 additional use cases next year…This is very promising.”
The Future: From Attended Automation to Full Autonomy
The event offered a window into the future of agentic automation. While many organizations start at lower levels of automation—like attended scenarios where humans initiate actions—the trajectory points toward fully autonomous, level-5 scenarios. In these advanced stages, automation not only executes tasks independently but also learns from outcomes, self-heals, and continuously refines its logic to handle emerging complexity.
Attendees were introduced to the concept of dynamic multi-agent systems. These systems involve specialized agents—some focusing on logic, some on decision-making, others on processing documents—working together without needing an extensively predefined workflow. Instead, the human sets high-level goals, and the system self-organizes to achieve them, adapting to changing conditions or entirely new tasks along the way.
This vision captivated audiences seeking to transcend simple process improvements. Instead, they look forward to a future where the automation layer becomes a responsive, adaptable partner, seamlessly coordinating tasks and optimizing results with minimal supervision.
Wrapping Up 2024 with Confidence
As the session concluded, the core message was empowering. Businesses can confidently approach agentic automation as a long-term strategy, evolving their use cases as they learn and build trust in these solutions. With the right approach, the future—one where human intervention is the exception rather than the norm—feels well within reach.
DeepOpinion’s 2024 End-of-Year Product Launch Event left attendees with a sense of possibility and direction. Far from a tech trend or marketing slogan, “Zero to Autonomy” emerged as a practical, achievable milestone, ushering in an era where back-office processes no longer hold organizations back but propel them forward.
.png)
