What if your predictive AI investments could start delivering agentic AI value now? According to DataRobot Chief Product Officer Venky Veeraraghavan and Dell Technologies Senior Director of AI Solutions Brad Maltz, they can. And now is the time to go after it.
Production models, clean data pipelines, optimization engines, and governance controls give agents the grounded business context they need to drive faster decisions and measurable outcomes. An orchestration and reasoning layer can connect these capabilities across teams, systems, and data silos, turning predictions into coordinated action.
In a recent DataRobot and Dell Technologies webinar, Veeraraghavan and Maltz explain how enterprises can build on the AI capabilities they already have and move quickly from predictive insights to agentic outcomes.
Agentic AI demos can make the technology feel magical: a chat interface appears to understand any request, navigate an entire workflow, and produce an answer. Inside the enterprise, the opportunity is practical and much closer than it appears.
Predictive AI already handles the hard analytical work. Models generate forecasts, scores, and recommendations within larger workflows that drive business outcomes. People interpret those outputs, consult dashboards, evaluate tradeoffs, run scenarios, coordinate across teams, and decide what happens next. Veeraraghavan calls this layer of interpretation and coordination “human middleware.”
As Veeraraghavan explains, the data and models at the center of these workflows provide the foundation for agentic AI. Agents connect that intelligence to the reasoning, coordination, and decision-making required to produce an outcome.
Agents accelerate the work surrounding the prediction. They interpret intent, break goals into smaller problems, call the appropriate data and analytical tools, synthesize the results, and surface a recommendation or exception to the person accountable for the outcome.
Language models provide flexible reasoning and orchestration. Enterprise data, predictive models, mathematical models, business rules, and optimization systems provide grounded, often deterministic answers. Combined in an agentic workflow, they create an adaptive path from business question to action.
Your existing AI investments already hold valuable intelligence. Agentic orchestration extends that intelligence across the decisions and actions that drive business results.
Agentic AI creates value at three levels, each with a different degree of impact, measurability, and strategic reach.
These tools help individuals create presentations, analyze information, write emails, and complete routine work faster. The productivity gain is real, but its financial impact can be difficult to quantify. Saving a few minutes on an email does not translate cleanly into revenue, margin, or reduced risk.
These agents accelerate established workflows inside platforms such as Salesforce, SAP, ServiceNow, and Workday. They can process expense reports, resolve service tickets, and complete other structured tasks more efficiently. Their impact is easier to measure, although it typically remains contained within one application, process, or function.
Agent workforces put agents at the center of consequential business workflows. They coordinate data, predictive models, optimization engines, applications, and human expertise around a defined outcome. Their impact can be measured through the business metrics leaders already track, including revenue, margin, operational efficiency, and risk.
Veeraraghavan connects this third category to the growing demand for demonstrable returns from enterprise AI investments. This is where existing predictive AI investments can compound.
Much of the analytical foundation may already be in place, including enterprise data, sensors, models, and optimization logic. Agentic orchestration connects those assets across the workflow, shortening the path from intelligence to decision to measurable business impact.
Chevron is applying agentic AI to a high-stakes challenge: protecting people during gas leaks and other anomalies at industrial facilities.
IoT sensors detect the anomaly. Models project how the gas plume will move under local weather conditions. An optimization engine directs tasks away from danger. An agentic application brings these capabilities together, allowing operators to evaluate scenarios and coordinate a response in near real time.
Speed matters. Electrical and mechanical drones can ignite leaking gas, while sending people into the affected area creates additional risk. Agentic orchestration gives operators a faster way to determine where the gas is moving, which equipment can operate safely, and how the response should adapt.
A technology company is applying the same pattern to supply-chain volatility. Quarterly forecasts and planning cycles could no longer keep pace with shifting demand, new technologies, logistics constraints, and changing customer priorities.
The company uses an agent to orchestrate its existing predictive models, what-if analysis, and optimization tools. A planner can evaluate what happens when inventory moves to another customer, compare delivery times and profit margins, and optimize for competing priorities such as meeting quarterly targets or protecting strategic accounts. Supply-chain and sales operations teams can then assess disruptions together and respond faster.
Both examples build on capabilities already in place: enterprise data, sensors, predictive models, and optimization logic. Agentic orchestration connects those assets in a responsive decision system, accelerating the path from signal to analysis to action.
Moving from predictive to agentic AI requires clear decisions about where to invest, how to architect the system, and which opportunities to pursue. These three principles can help enterprises focus resources on measurable value while building the flexibility to evolve.
A cost strategy should start with two questions: What should run, and where should it run?
The answer may combine frontier and open-weight models across cloud, on-premises, deskside, and edge infrastructure. A complex reasoning task may justify a frontier model, while a smaller open-weight model may handle a simple, repetitive step more efficiently. Data sensitivity, latency, quality, and control all shape the economics. Maltz noted that, for some workloads, on-premises or deskside approaches can reach break-even against hosted environments within months.
The model landscape is changing too quickly to make one provider or model the permanent answer to every task. Treat models as a portfolio. Route each request according to the criteria that matter for that step, including quality, cost, latency, data sensitivity, and deployment requirements.
This approach also keeps the architecture open to improvement. A predictive model can remain a tool the agent calls today and be replaced later when a better option emerges. The workflow continues delivering value as its individual components evolve.
The highest-value opportunities often span several teams, systems, and data silos. Consider outcomes such as responding to a supply-chain disruption, completing a know-your-customer review, protecting plant safety, or optimizing a tariff decision.
Work backward from the outcome. What data grounds the agent? Which models, applications, and business rules must it call? What actions can it take? Where should a subject-matter expert approve, intervene, or handle an exception? These questions reveal where agentic orchestration can produce meaningful business impact.
Enterprise readiness for agentic AI already exists across production models, governed data, domain expertise, business applications, infrastructure, and years of operational learning. Connecting these capabilities around a high-value outcome creates a practical path forward.
Start with predictive systems you trust and make them available to agents as tools. Add controls, observability, and human oversight. Measure performance through business outcomes, then improve individual components as the workflow evolves.
DataRobot’s recognition as a Leader in the Gartner® Magic Quadrant™ for Data Science and Machine Learning Platforms for the third consecutive year reinforces the maturity of this foundation. Production-grade agentic AI is ready to move from experimentation into consequential business workflows.
Enterprises with useful predictive models and trustworthy data may already have the foundation they need. Agentic orchestration can turn those investments into coordinated action and measurable value.
Watch the full DataRobot and Dell Technologies webinar to learn how enterprises can build on their predictive AI investments and move toward agentic workflows that deliver measurable business value.
The post Your predictive AI foundation is the fastest path to agentic AI value appeared first on DataRobot.
Image credit: Olivier Porchet, Biorobotics Laboratory, EPFL
When a fish holds its position against a current in a river, its brain must figure out how fast to swim and how to steer to offset the water flow. Most fish use vision to register the world sliding past, detect optic flow speed and direction, and their brains turn these signals into compensatory swimming. Neuroscientists call this stabilizing reflex the optomotor response (OMR), from neural activity imaging. The retina captures signals of optic-flow direction, central pretectal neurons interpret direction, and spinal nerves drive muscle contractions.
The catch is that one cannot easily change the living brain to test how these circuits work. Although advances in imaging now allow detailed recording, and even manipulation, of neurons alongside behavior, rewiring connections to ask what a particular link actually does remains almost impossible in the living, complex animal.
A joint team from EPFL (Switzerland), Duke University (USA), and the Instituto Superior Técnico (Portugal) took on this challenge by creating a realistic larval zebrafish simulation and a biomimetic robot, published in Science Robotics. The team found a way to replicate the body and known neural circuit architectures of live zebrafish, first as a physics-based simulation, then as a free-swimming robot that autonomously navigates upstream using vision and these bio-inspired neural circuits.
The results revealed the minimal set of neural components needed for the OMR and, more interestingly, showed that this balanced neural circuit allows autonomous upstream navigation even in poor visibility. It also highlighted that the fish’s body and eyes are part of the neural computation.
A blueprint from the living fish brain
The starting point was work in the Naumann Lab at Duke, where detailed behavioral studies and whole-brain calcium imaging of larval zebrafish exposed to visual stimuli that mimicked riverbed optic flow, paired with circuit modeling, yielded a best-fit wiring diagram of the brain-scale OMR pathways.
Drawing on other insights about neural processing in the vertebrate retina and spinal cord, Dr. Xiangxiao Liu and Luca Zunino from the EPFL team used this neural circuit model to develop a neuromechanical simulation, simZFish, that not only recreates the larval fish body, complete with eyes and fins, but also takes this experimentally derived brain blueprint to be the simulation’s brain, opening new paths for neuroscience and brain-inspired robotics.
simZFish: a brain you can take apart
Built in the physics-based Webots simulator, simZFish reproduces a six-day-old larva at 1:1 scale: a tiny 4 mm body, weighing only 0.3 mg with seven segments, driven by six simulated motors, a head with two sideways-facing cameras for eyes, and realistic water fluid dynamics. With just the right head-to-tail weight balance, the simulated simZFish moves just like real larval zebrafish. Its artificial brain replicates the entire neural circuit found in fish, from light-changing pixels to muscle activation.
The pretectum is a visual brain region that contains neurons that receive direct input from the retina, computing motion directions. The artificial retina detects motion and feeds four types of direction-selective ganglion cells, which drive pretectal neurons that integrate and process visual information from both eyes, and downstream hindbrain motor command neurons that set how often the fish swims and which way it turns. Finally, to emulate how the real fish swims in intermittent bouts, a “bout gate” releases a burst of tail beats, producing the characteristic burst-and-glide swimming of real larval zebrafish.
Because every part is simulated, researchers can change any aspect of simZFish’s body or neural connection weights, delete or add neurons, or change the eye’s lens and see the consequences immediately. In contrast, with animal experiments, one can only record correlations of neural activation if the fish happens to execute the behavior in question, but cannot exclude that some other processing was going on or easily change or interact with the internal structure. simZFish turns that black box into an open, well-lit one with identified components, allowing the team to pinpoint the “minimal essential elements” for OMR behavior.
Figure 1. The larval zebrafish-sized simZFish can swim around in simulated water in a virtual Petri dish and be presented with an unlimited number of visual environments. SimZFish has two laterally placed virtual cameras that can ‘see’ the virtual environment (top-left boxes), a sensorimotor controller, and six motors linked in series in the tail, enabling it to capture and respond to visual motion information.
The body is part of the computation
One specific insight from building this bio-inspired system concerned retinal-brain connectivity. With cameras on the sides of simZFish’s head, optic flow generated when fish are dragged in a river produces conflicting swirls across the visual fields, highlighting a version of the classic “aperture problem”. Therefore, feeding the motion information from the entire simulated retina into the circuit caused those signals to cancel out, breaking the OMR behavior. When the team restricted input to the lower posterior part of the visual field, the behavior snapped back into place.
Strikingly, that is exactly the region that most strongly drives the OMR in real zebrafish, and it matches the large, lower-posterior receptive fields neuroscientists have recorded in the real fish’s pretectum. The insight is not so much that the simulation “reveals” the circuit, but that embodiment, in this case the perspective distortion of the laterally placed eyes, explains why the circuit is wired the way it is: the layout of the body and eyes dictates a configuration that captures the most useful motion information with the fewest connections, an efficient solution evolution appears to have found as well.
Closing the loop: a prediction sends the neuroscientists back to the microscope
The simulation also predicted that the original neural circuit model was incomplete because it could not readily test behavioral responses to new stimuli. However, when each simulated eye was shown motion in the opposite direction, a shearing stimulus the model had never encountered, simZFish1.0 turned far more than expected. When our lead neurobiologist, Dr. Matthew Loring, showed these visual stimuli to real zebrafish, they barely turned.
That mismatch sent the neuroscientists at Duke back to the microscope. Using volumetric two-photon calcium imaging, they recorded tens of thousands of neurons. They found that certain binocular neurons in the pretectum, specific response types that the circuit model relies on, existed in multiple previously overlooked subtypes. The team updated the too-simplistic simZFish1.0 model using neurons predicted by the simulation, which come in forward- and backward-tuned subtypes, with responses to one eye suppressed when the other eye sees forward motion.
Adding these subtypes and updating the connectivity produced simZFish 2.0, which reproduced the real fish’s behavior far better, including the response to the artificial shearing stimulus. “simZFish narrows the search, and the animal experiments verify the predictions, and together they complete the puzzle of the neural circuit,” said EPFL’s Auke Ijspeert, who directs the Biorobotics Laboratory (BioRob) at EPFL.
The final test came when the team released simZFish into a simulated flowing river, dragging its body with realistic forces, now with a realistic riverbed rather than artificial moving stripes. Strikingly, no matter which direction the simZFish started in, it eventually turned and swam upstream. “It was remarkable to observe that our experimentally derived neural circuits were allowing the simulation to navigate upstream autonomously”, marveled Eva Naumann, Assistant Professor of Neurobiology and Secondary Biomedical Engineering at Duke University.
Into the wild: a robot that keeps up with the current using vision alone
To see whether a neural circuit tuned in a clean simulation could cope with the real world, the team scaled the tiny simZFish design up into ZBot, an 80-centimeter, 2.7-kilogram robotic implementation carrying two real cameras, six tail motors, and a Raspberry Pi running the very same neural network, in a water-tight enclosed head.
Dropped into the sometimes-muddy, turbulent Chamberonne river near Lausanne, with plenty of river rocks, dappled light, drifting leaves, ZBot used the OMR circuit to counteract the water flow. With its visually activated OMR circuit, it stayed in the aerial drone camera’s view for much longer, 58 seconds, compared with roughly 37 seconds when its “eyes” were switched off and just 20 seconds when it drifted with the motors off (Figure 2, Video 1).
Figure 2. ZBot swimming in the Chamberonne River, Canton de Vaud, Switzerland.
It is the first demonstration that a fish could, in principle, fight a current using visual circuits alone. This overturns a long-held assumption that this behavior, rheotaxis, requires the mechanical “lateral line” network of sensory organs along the side of the fish body to detect water flow. “This is the first proof that effective rheotaxis can be achieved using visual neural circuits alone,” noted Professor Ijspeert.
Video 1. With the OMR neural model embodied, ZBot maintains its position far better than with random bouting (OMR circuit blinded) or passive drifting (motors off).
Why it matters for robotics
For robotics, the appeal is building more efficient, more autonomous agents. Most drones employ dedicated downward-facing cameras for position stabilization. By contrast, our method leverages the existing lateral cameras and requires no additional hardware, cutting both hardware and computational costs. Interestingly, this balancing circuit that compares across eyes seemed to handle poor visibility very well, likely also because the burst-and-glide swimming allowed the ZBot to integrate useful sensory information during the glide phase, when visual input was not degraded by self-motion.
The same bio-inspired burst-and-glide control of the ZBot underlies a companion study by the team showing that intermittent swimming saves energy, pointing toward lighter, longer-lasting aquatic robots that switch gaits as real fish do. Both the simulator and the robot designs are open-source, so other groups can reuse the models or build and test new circuits.
Yet, as the ZBot is roughly 200 times larger than its biological model and swims in a different fluid regime, it tests the circuit’s logic rather than the larva’s exact mechanics. The realistic simZFish’s visual circuit also holds its position only against a gentle flow, because the model currently cannot dynamically respond to different flow regimes. For the ZBot, about half the river trials were discarded for collisions or signal dropouts, probably because the stripped-down OMR circuit doesn’t allow the ZBot to avoid obstacles.
Even so, the approach the team built offers a powerful template for connecting brains, bodies, and behavior. “By integrating simulation, robotics, and animal experiments, we get a far more complete view of an animal’s neural model,” said Naumann, who runs a systems neuroscience lab testing how visual neural circuits interact in live zebrafish. Starting from these tiny baby fish, this iterative simulation-neurobiology-robotics approach is opening new ways to understand animal intelligence and to build efficient, brain-inspired aquatic robots.
Paper and team information
You can access the simZFish open-source platform here and explore more work on the Naumann Lab Github.
EPFL led simZFish development, neuromechanical modeling, and ZBot design and testing. Duke University led the zebrafish behavioral experiments, two-photon calcium imaging, and neural network modeling and data analysis. The Instituto Superior Técnico contributed modeling and theoretical analysis of the visual neuromechanical system.
Artificial Intelligence (AI) has emerged as a transformative force, enabling video streaming platforms to deliver personalized experiences, enhance content discovery, and drive user engagement. As the media and entertainment industry continues to expand, integrating AI into video streaming services has become essential for remaining competitive and meeting the growing demands of viewers.
Harness next-gen streaming with AI, where growth and user engagement meet. AI technologies such as Machine Learning and Natural Language Processing have revolutionized the dynamics of how streaming services interact with users. Through big data set analysis, AI provides customized recommendations for content that increase user satisfaction and retention. The top video streaming applications, including Netflix and YouTube, leverage AI to suggest videos based on watch history, behavior, and interests, leading to increased watch time and user engagement.
Furthermore, AI-powered features such as voice search, auto-subtitles, and live language translation have enhanced the availability of content to a broader international audience. These innovations not only enhance the user experience but also expand the reach of streaming services to diverse demographics.
AI Integration into video streaming apps and platforms has been at the forefront. According to a Grand View Research report, the video streaming market globally is projected to grow to USD 416.84 billion by the year 2030 with a compound annual growth rate (CAGR) of 21.5% between 2025 and 2030.
Increasing AI adoption is the major driver that improve the consumption of content, optimize advertising, and enhance user engagement.
AI can also contribute significantly to content creation and curation. It automates processes such as video editing and metadata tagging, accelerating automated content creation and reducing the operational costs. Additionally, AI analytics provide insights into viewers’ interests and actions, enabling platforms to personalize their content strategies.
AI is transforming how viewers interact with video content. From personalized recommendations and real-time subtitles to predictive streaming and immersive AR/VR, these features boost engagement, retention, and revenue. Key features include:
Recommended To Read: Best Video Streaming Apps in the USA
The AI-powered video streaming app development would require an initial investment of $80,000 to $250,000+. AI features like personalized recommendations, dynamic ad targeting, predictive bandwidth optimization, and real-time content moderation not only enhance user engagement but also increase subscription retention and ad revenue.
The cost of developing an AI-powered video streaming app depends on several factors, including the complexity of AI features, platform choice (iOS, Android, or Web), design, backend infrastructure, and scalability requirements. Key cost drivers include:
Partnering with an experienced AI development company like USM Business Systems ensures cost-effective solutions that balance advanced features with performance and scalability, delivering maximum ROI.
AI is no longer a luxury but a necessity in the video streaming industry. By harnessing the power of AI, platforms can offer personalized experiences, optimize operations, and drive growth. For businesses looking to stay ahead in this competitive landscape, integrating AI into their video streaming services is a strategic move that promises substantial returns.
For businesses in the video streaming industry, partnering with AI development experts can unlock new opportunities for innovation and growth. AI solutions can be customized to address specific business needs, whether it’s enhancing content discovery, improving user engagement, or optimizing monetization strategies.
At USM Business Systems, we specialize in developing AI-driven solutions tailored for the video streaming industry. Our expertise in machine learning, data analytics, and cloud technologies enables us to create scalable and efficient AI applications that drive business success.
OpenAI has slashed the price of one of its most powerful AI engines –GPT-5.6 Luna – by 80%.
The move is seen by many as OpenAI’s attempt to compete with Chinese-made AI, which is viewed as nearly as good as U.S. AI – and often available for pennies-on-the-dollar.
The reduced price for Luna is available to users who access OpenAI’s servers directly from their computers – rather than to users who work with AI via ChatGPT.
In other news and analysis on AI writing:
*China’s Nearly as Good AI – Kimi 3 – Now Available for Free Download: Researchers and other organizations with gigantic computer systems can now download OpenSource software Kimi 3 and experiment with it freely.
That ability to play with and customize Kimi 3 at no charge helps keep the pricing down on AI from companies like OpenAI and Anthropic, which don’t offer such free access to researchers.
Observes Bloomberg News: “Made-in-China AI models have surged in global popularity in recent times — led by DeepSeek — as they offer more affordable rates and comparable performance to U.S. offerings.”
*Amazon: The Pay-Off for AI Investment is Real: Amazon’s decision to invest heavily in building-out computer infrastructure to support AI in the cloud just brought back the bacon.
Observes writer Nathan Bomey: “Amazon’s AI cloud business soared in the second quarter, driving far higher revenue than expected.
“Amazon’s North America sales rose 16%, international sales increased 15% and AWS sales jumped 37%.”
*Microsoft: The Pay-Off for AI Investment is Real: Microsoft profits jumped 31% in Q2 for 2026, largely credited to its heavy investment in AI infrastructure.
Observes writer Emmy Martin: “Revenue at Azure, the company’s cloud computing service — which lets businesses rent computer power, storage and AI tools — drove much of the growth, surging 43%.
“The company said more than 30 million people were paying for Copilot, up from more than 20 million a quarter earlier.”
*AI Law Firm Will Insure Legal Work By Its AI Agents: AI-powered law firm Crosby is promising to include legal liability insurance for all work performed by its AI agents.
Crosby’s pitch: The firm’s AI agents will be held to the same liability standard as a human lawyer.
Observes Artificial Lawyer: “This is all about the autonomy of the legal agent — of trusting the agent to do the work.”
*ChatGPT Adds Yelp to its Database: Answers to some of your ChatGPT questions will now also rely on data supplied by business review platform Yelp.
That change should especially come in handy when you’re researching local businesses and services using ChatGPT and are looking for business reviews, photos and similar information currently stored on Yelp.
Observes writer Madison Mills: “As AI becomes a new front door to the Internet, content companies (like Yelp) are deciding whether to partner with that ecosystem — or compete with it.”
*Oops: AI Agent Triggers Million-Dollar-Plus Cost Overrun: Amazon found out the hard way that using a ‘set-it-and-forget-it’ AI agent can cost you. Big time.
Specifically, Amazon says one of its AI-agent-driven projects went over-budget by 860%, triggering cost-overruns at $1.8 million.
Observes writer Jowi Morales: “These mistakes used to be trivially cheap, but AI models made them catastrophically expensive — especially as token spending drastically increased with the deployment of AI agents.”
*ChatGPT Will No Longer Write Like Famous Authors: One of the fun perks of ChatGPT – getting it to write in the style of your favorite author – is disappearing.
Instead, ChatGPT offers to incorporate the “broad qualities” of the writing style of say Stephen King, Charles Dickens or Ernest Hemingway, according to writer Kyle Orland.
Adds Orland: “Not all major LLMs (AI engines) treat these style imitation requests the same way, though. No Latency’s study found that Google’s Gemini consistently complied with requests to copy an author’s style.”
*Summer 2026 Guide to AI: One Prof’s Winners: Wharton Professor and AI expert Ethan Mollick has just released his guide to the best in AI.
His overall advice: For important work, don’t penny pinch. Use the best AI money can buy.
Observes Mollick: “For these issues, you will want to use the most advanced models you can get access to, which is either Claude’s most powerful models, Opus and Fable, or ChatGPT’s GPT-5.6 Sol, set to at least the “High” thinking levels.
“That is because these models have lower error rates and score much higher on ability tests in complex fields.”
*EPA: Feel Free to Pollute Our Environment, AI: The U.S. EPA has ruled that electricity power plants built exclusively to supply AI data centers are exempt from the Clean Air Act’s regulations on acid rain.
Observes Reuters: “The move is aimed at accelerating AI infrastructure development, easing pressure on regional power grids, and supporting the rapid expansion of data centers across the United States.”
Acid rain has been shown to trigger aquatic ecosystem collapse and degrade forests and soil.
Share a Link: Please consider sharing a link to https://RobotWritersAI.com from your blog, social media post, publication or emails. More links leading to RobotWritersAI.com helps everyone interested in AI-generated writing.
–Joe Dysart is editor of RobotWritersAI.com and a tech journalist with 20+ years experience. His work has appeared in 150+ publications, including The New York Times and the Financial Times of London.
The post ChatGPT-Maker Slashes Some AI Prices by 80% appeared first on Robot Writers AI.