Shopflo claims D2C brands do not need their own AI team for growth and discovery

For companies implementing artificial intelligence, key questions are not only about choosing a model but also about where the AI operates, its cost, and the ability to keep confidential data under the company's control. PeakVision AI offers a solution that involves moving most of the AI tasks inside the enterprise.

The company's system considers the task, rather than an individual application, as the main unit of work: users state a goal, and AI agents perform the necessary steps in a local computing environment. Co-founder and CTO Liangyu Nie told TechNode that PeakVision AI was founded in 2024, overcoming two major obstacles to AI adoption in corporations: the complexity of technology deployment and concerns over data security.

In March 2026, the company opened its headquarters in Hangzhou and received government support for research and development amounting to 15 million yuan. In May, it presented its first enterprise systems and completed an initial funding round, followed by a second round worth tens of millions of yuan in August.

PeakVision AI creates an enterprise AI stack that integrates locally deployed computing, FengKan AIOS, and AI agent services. This system is designed to focus on how agents complete tasks, rather than how people move between different applications. As Nie noted, users no longer need to load different applications; they just need to tell the AI what they want via a chat window, and it will do the work.

The startup's central thesis is that enterprise software will increasingly be structured around intent and execution, rather than menus, files, and separate applications. The company plans to make FengKan AIOS open source by the end of October. PeakVision AI monetizes through hardware, subscriptions, and custom enterprise AI implementations. The company's broader goal is to transform project-specific work into universal solutions for various industries.

FengKan AIOS supports third-party hardware, while PeakVision AI takes a selective approach to its skills ecosystem. Nie clarified that users' end skills, which may contain proprietary workflows, will remain untouched. However, more generalized skills created by partners and leading engineers can be released to the market.

Currently, the company is testing its model through the FDE academy, launched by Shanghai Sixunited Intelligent Technology with support from Advanced Micro Devices (AMD). AMD provides computing hardware, technical toolchains, and resources for projects. PeakVision AI also collaborates with chip manufacturers in early product development stages to optimize costs and performance, including a partnership with ChangXin Memory Technologies to reduce memory costs.

Unlike companies such as Wonderful, which target large, often regulated organizations and provide them with AI platforms and implementation teams, PeakVision AI targets small and medium-sized businesses. Nie emphasized that these enterprises face the same problems—costs, technical complexity, and data security—but have fewer resources to solve them. PeakVision AI uses inexpensive hardware from the Chinese supply chain so that each implementation becomes a repeatable solution for similar clients.

Nie stated:

The future of technology points towards the development of agentic artificial intelligence, and the adoption of this technology is gaining momentum. As part of the expanded partnership between Cognizant and Google Cloud, the 'Gemini Frontier Capability Group' has been formed. This group combines Google's ready-to-use AI platform with Cognizant's industry and engineering expertise as an AI developer.

This group is supported by over 18,000 Cognizant employees and significant joint investments. Concurrently, ahead of its Initial Public Offering (IPO), the anchor package of Jio Platforms attracted bids totaling 2.5 lakh rupees. The company allocated 7.24 million shares at a price of 1,119 rupees to anchor shareholders, with LIC receiving the largest stake.

Furthermore, the startup Hunar.AI raised 40 crore rupees in a Series A funding round led by Blume Ventures. Hyderabad Angels Fund, Together Fund, Seedstars, AVNM Ventures, co-founders Prateek Maheshwari and Alakh Pandey of PhysicsWallah, and co-founder Aditya Wuchi of MediaMint also participated in the round. This startup, which develops specialized AI technologies for front-line intensive industries, plans to use the funds to build AI infrastructure, develop products, and expand sales channels.

According to a Redseer report, the country's online retail market is expected to exceed $90 billion this year, with grocery products surpassing mobile devices. E-commerce volume grew by approximately 25% year-over-year, reaching about $38 billion in the first half of 2026, marking the strongest first-half performance in four years.

This briefing examines Amazon's bet on agentic AI for enterprise operations, as well as the question of marketing AI tool effectiveness: do they truly increase profit or are they just noise?

Technology Event Overview

At the TechSparks 2026 event, Rajiv Rastogi, Vice President of Machine Learning at Amazon, emphasized the growing role of agentic AI across all layers of enterprise operations. He noted the shift from static software to autonomous systems capable of understanding natural language instructions.

Rastogi explained during a fireside chat at the 17th edition of the flagship tech startup event YourStory that 'agents dynamically prepare plans to solve problems and invoke relevant tools.'

Driving Force of Transformation

These autonomous agents can evaluate requests, formulate a sequence of actions, and call upon specialized digital tools to complete complex tasks end-to-end, rather than relying on human intervention at every stage. Within Amazon, this operational layer functions in systems like Rufus—a conversational shopping assistant that analyzes customer queries, applies price filters, calls search engines, and guides users through product comparison and checkout.

Similarly, autonomous seller assistants pre-fill product descriptions, monitor inventory, and help small businesses comply with international export requirements.

In Rastogi's view, the industry will ultimately move towards true global models that possess an intrinsic understanding of physical spaces and environmental behavior, creating a practical and economically viable foundation for business automation.

Marketing and Decision Making

At TechSparks 2026, Anand Jain, Co-founder and CMO of CleverTap, a marketing automation and customer engagement platform, elaborated on why marketing teams must move beyond basic generative AI productivity tools and transition to real-time autonomous decision-making systems that deliver tangible business profit.

During a marketing strategy masterclass, Jain urged leaders not to limit themselves to simple gimmicks but to move from using AI for entertainment to achieving actual business outcomes. He questioned whether AI-generated content creation truly contributes to growth, asking, 'Will this actually impact business results, or is it just the same thing?'

Instead of broadcasting generic promotional offers to wide audiences, modern marketing mechanisms must interpret customer actions in real time to meet individual needs. This approach signifies a transition to the highest level of AI maturity, known as 'real-time decision-making.'

The news and updates section noted the entry into force of a free trade agreement between India and New Zealand, which grants duty-free access to a range of Indian goods, including products from labor-intensive sectors such as textiles and leather. It was also reported that OYO's parent company, PRISM, plans to nearly halve the size of the new offering in the proposed IPO, reducing it from almost 6,650 crore to 3,350 crore rupees, due to strong cash generation and balance sheet improvement. Furthermore, US Secretary of State Marco Rubio arrived in India after President Donald Trump postponed US tariffs on imports from India, as both countries work to finalize a trade agreement. During his two-day visit to Mumbai and Jamnagar, Rubio will meet with government and business leaders.

Most modern marketing teams use artificial intelligence to save time on tasks such as drafting emails, creating image variations, and summarizing long documents. However, Anand Jain, co-founder and CMO of CleverTap, a platform for marketing automation and customer engagement, argues that basic tools performing these functions rarely contribute to increased company profit.

During a marketing strategy masterclass at the TechSparks 2026 event, Jain urged leaders to look beyond simple ways to speed up work. He emphasized the need to shift from using AI for entertainment to applying it for real business results.

While AI-powered content creation speeds up work and saves time, Jain questioned whether this truly contributes to growth, asking: 'Will this actually impact business outcomes, or is it just the same thing?'

Traditional digital campaigns usually operate on a 'push' model, where companies decide what they want to sell instead of understanding what consumers want to buy. In contrast, an offline consultant remembers a client's personal preferences, notices instant reactions, and suggests products matching their budget and style. Digital marketing must adopt a similar approach.

Jain stated: 'Not what I want to sell, but what my customer wants to buy. That's how it needs to be flipped.' Instead of blasting general promotions to wide audiences, modern marketing systems should analyze customer actions in real time to match individual needs.

This method marks a transition to the highest level of AI maturity, known as 'real-time decision making.' Marketing technologies typically evolve through three stages. The first level focuses on basic time-saving tasks, the second covers experiments like A/B testing (comparing two ad versions to determine the best), and the third achieves true automated decision-making.

Jain noted: 'You start with productivity, then you run experiments, and the hardest part is decision-making. When you are actually building an entire system for one group of people, and that one person could be one of several millions appearing in your app or on your website.'

Instead of bombarding users with promo codes, a decision-making system first offers soft, helpful steps, such as building brand trust or explaining product benefits. Real-time decision-making relies on a single data hub that combines the historical customer profile with their recent real-time actions. Traditional A/B testing usually stops after selecting one fixed winner, whereas a decision-making system uses reinforcement learning—a smart system that continuously learns from user responses and adjusts offers on the fly.

Furthermore, instead of relying solely on propensity scoring (calculating the probability of purchase), a decision-making system uses uplift modeling. This method identifies the specific nudge, message, or channel needed to prompt an individual user to take the next positive step.

As automated systems manage campaign delivery across thousands of message variations, the marketer's role shifts from execution to strategic guidance. Instead of manually creating creative variations, marketers define high-level goals and set strict operational constraints. Jain emphasized: 'Your job is to create budget constraints, brand constraints, regulatory constraints. Do not use the word 'free' if you are offering something, if you are a bank.'

By setting clear boundaries, marketing leaders allow intelligent decision-making systems to operate autonomously, maintaining the safety of brand messaging while simultaneously ensuring genuine commercial profitability, he added.