In this exclusive interview, Rohan Manthani, Co-founder, Streamoid, shares with Anannya Saraswat, Reporter (Public Sector & Leadership) at CXO Media and APAC Media, how domain-focused AI (AI models trained for specific datasets for a particular industry) is reshaping retail and fashion by moving beyond experimentation to real operational impact.
He shares how Streamoid uses generative AI to reduce inefficiencies across design, cataloguing, and planning in the retail sector. The discussion also looks at future AI-native platforms, evolving retail skills, and the next big shifts shaping an AI-first retail ecosystem.
Many believe vertical AI will outperform general models in real-world deployment. How do you see domain-focused models transforming industries like retail and fashion?
General AI models are powerful, but they require significant context to perform reliably in real-world processes. In fashion and retail, decisions depend on highly specific factors such as seasonality, regional preferences, product attributes and historical performance. Domain-focused AI models are better suited to understanding these nuances.
In practice, this allows AI to move beyond surface-level generation and support real operational decisions. When models are trained and fine-tuned on retail-specific data and workflows, they can help teams decide what to design, how to describe products, how much to produce, and where to allocate inventory. This is where vertical AI becomes transformative, not as a tool for experimentation, but as a system that supports daily execution.
Gen-AI is often considered a creativity booster. How is it solving hard operational problems and removing inefficiencies in retail workflows?
While generative AI is widely associated with creative use cases, its biggest impact in retail is operational. Many retail workflows remain manual, fragmented, and spreadsheet-driven, especially across cataloguing, planning and cross-team coordination.
GenAI helps by automating repetitive tasks such as product enrichment, content creation, and data standardisation. More importantly, when AI is embedded directly into workflows, it connects design, cataloguing, and planning, so decisions are not made in isolation. This reduces handoffs, errors, and delays, enabling teams to move faster with fewer resources while maintaining quality.
For enterprises, what are the biggest blockers to scaling AI beyond pilots?
The biggest blocker is fragmentation. Many enterprises experiment with AI through isolated pilots that are not connected to core systems or workflows. As a result, value remains limited and difficult to scale.
Data quality is another major challenge. In retail, inconsistent and unstructured product data makes AI outputs unreliable. Without a single source of truth, teams struggle to trust results. Scaling AI requires unified data, clear workflows, and systems designed for operational use rather than experimentation.
Fashion is driven by design sensibilities. How does Streamoid ensure AI recommendations feel human and on-trend?
AI does not replace creative judgment. At Streamoid, AI is designed to surface signals rather than dictate outcomes. Recommendations are grounded in brand guidelines, historical performance, and market context, ensuring they remain aligned with each brand’s identity.
Designers retain full control over final decisions. AI narrows the range of options and highlights what is likely to resonate, allowing creative teams to focus more on refinement, storytelling, and originality while staying informed by real demand.
AI needs strong data foundations. What challenges do you see in data standardisation and infrastructure in the retail sector?
Retail data is often incomplete, inconsistent, and spread across multiple systems. Product attributes, images, descriptions, and performance data are rarely standardised, which limits how effectively AI can be applied.
This lack of structure slows automation and increases manual work. Addressing this requires platforms that unify data across the product lifecycle and continuously improve data quality. Strong data foundations are essential for AI to move from insights to reliable execution.
With generative AI tools becoming common, what are the differentiating factors that make Streamoid unique?
Many generative AI tools operate as standalone solutions. Streamoid is built as an AI-native platform where design, cataloguing, planning, and commerce are interconnected through a shared data and intelligence layer.
Instead of teams prompting models manually, Streamoid orchestrates context automatically using brand and operational data. This allows AI to deliver accurate, consistent outputs that improve with use. The focus is not on isolated generation, but on enabling end-to-end workflows that drive real business outcomes.
From catalogue data to customer experience and supply chain visibility, where do you see the next biggest ROI for AI in retail?
The biggest ROI comes from connecting decisions across the lifecycle. Improving catalogue data enhances discovery and conversion, but the real impact emerges when this intelligence feeds into planning and production decisions.
Better demand visibility reduces overproduction, improves inventory allocation, and shortens time to market. When AI connects what customers buy with what brands design and produce next, efficiency gains compound across functions rather than remaining siloed.
Streamoid operates at the intersection of tech and retail/D2C. What future skills will retail teams need to thrive in an AI-first workplace?
Retail teams will increasingly need the ability to work alongside AI systems. This includes interpreting AI-driven insights, providing clear direction, and applying judgement rather than executing repetitive tasks.
Creative, strategic, and cross-functional skills will become more important, while manual coordination work will decline. The most successful teams will combine deep domain expertise with the ability to use AI as a daily operating tool.
Looking ahead, what bold AI shifts do you think will reshape global retail?
The next major shift will be the move from fragmented tools to unified, agent-based platforms. AI will become embedded into daily workflows rather than used only for experimentation or isolated tasks.
Retail operations will increasingly rely on closed-loop systems where outcomes continuously inform future decisions. This will allow brands to operate with greater speed, consistency, and confidence, even as complexity increases.
