Ramco’s Launches Chia Targets 60% Ticket Deflection in Enterprise AI Push

Ramco Systems has officially entered the agentic AI market with the launch of Chia, an enterprise-grade conversational AI platform that the company claims can deflect up to 50%–60% of customer support tickets while operating with near-zero hallucination rates.

The platform launched on Thursday is a no-code conversational AI agent platform that enables customer service teams to automate complex, multi-step support workflows without engineering resources, executing actions across enterprise systems while maintaining strict compliance and governance controls, Ramco said.

The Chennai-based enterprise software provider is entering a conversational AI market valued at $12 to $14 billion globally and growing at 20% to 25% annually, according to Managing Director Abinav Raja. The product was tested internally for two quarters, and it represents the first of “many more” AI-native products planned for Ramco's portfolio, he said.

How Chia Works

What differentiates Chia from traditional chatbots and AI assistants, according to Ramco executives, is its deterministic approach to AI workflows. The platform uses what the company calls “Natural Language Workflow” (NLW) capabilities, allowing non-technical teams to define AI behavior in plain English.

“Non-technical people can set this up and modify it whenever they want,” Easwar Arumugam, Product Head of Chia, said, explaining that tasks which traditionally required extensive coding and backend engineering can now be handled without writing code.

The system employs a multi-agent architecture in which each agent performs clearly defined tasks within controlled boundaries, with structured orchestration reducing the likelihood of hallucinations or inaccurate responses.

During a live demonstration on WhatsApp, the system handled a complex product return scenario, including image verification of damaged goods, policy validation, and payment processing, all without human intervention. However, human intervention can be configured within the workflow if required, the company said.

Chia can integrate across Ramco's existing ERP products, including payroll, aviation, and HCM solutions, while also being sold as a standalone product.

Deployment in Weeks, Not Months

On deployment, the company said that while many AI agent pilots take several months to roll out, Chia can be deployed within weeks depending on enterprise readiness. Easwar Arumugam, Product Head of Chia, said individual use cases can go live in as little as two weeks if data is readily available. If organizations need to prepare documentation and APIs, deployment may take four to six weeks, he said.

Chia supports over 100 languages and can handle mixed-language conversations. Voice capabilities are slated for release in about a month, with video support also on the roadmap, the executives said.

Addressing enterprise security concerns, Raja emphasized that all models are hosted on private cloud instances. “All of these models are hosted on cloud providers, data is hosted locally in that particular geography region,” he said. The data never leaves the geographic region and is never used for training foundation models, with private hosting available on AWS, Azure, and Google Cloud.

From Internal Tool to Market Product

The genesis of Chia came from Ramco's own struggles with customer support, Raja revealed. “This was started off because we were solving this internally for our own use cases,” he said. “We saw that we were able to handle a lot of deflections through this, customer satisfaction was better, productivity improved because of this. So, we felt we should definitely take this out in the market.”

The company said it tested the product internally for over two quarters, during which it was able to deflect 50% to 60% of queries.

Ramco is positioning Chia not as a cost-cutting tool but as a revenue driver. “We want to position this as an opportunity for enterprises to actually increase revenue with a product like this, and not see it purely as an opportunity to reduce people costs,” Raja said.

President & COO Sandesh Bilagi added, “Customers today expect accuracy, speed, and seamless support across every touchpoint. Chia rises to this challenge by enabling enterprises to automate complex customer interactions with confidence and control.”

Pricing Model

Breaking from traditional SaaS pricing, Chia uses an outcome-based model. “There is no upfront cost. There is no seat pricing. Purely outcome-based,” Raja stated. Companies pay only when Chia successfully resolves a customer query without human escalation.

“In India, as far as I know, we are the first ones to adopt outcome-based pricing in this space,” Raja claimed, though he acknowledged that some global competitors have similar models.

Ramco is targeting enterprise customers across BFSI (banking, financial services, insurance), telecom, e-commerce, retail, healthcare, and aviation sectors. The geographic focus includes India, Asia, Australia, New Zealand, and the United States—markets where Ramco already has a strong presence.

The immediate go-to-market strategy is to prioritize existing Ramco customers, Raja said, referring to the company's plan to cross-sell Chia across its customer base of over 800 global clients.

When asked why companies would pay for a SaaS platform instead of building similar tools themselves, Raja said businesses are paying for reliability at scale. Deploying AI across thousands of customers while ensuring data security, privacy, and performance requires a structured enterprise platform.

He also addressed concerns that AI tools could disrupt the SaaS model, saying enterprise software companies will continue to play a key role because AI remains non-deterministic. Since it cannot be fully trusted to run an entire organisation on its own, businesses still need controlled, secure platforms, he said.

Looking Ahead

As organizations grapple with the promise and challenges of AI adoption, Ramco's approach with Chia represents a bet on controlled, deterministic AI that operates within enterprise guardrails. The 50%–60% ticket deflection rate from internal testing will now face real-world validation as the platform rolls out to enterprise customers in the coming months. The launch of Chia marks an early step in Ramco’s broader transition toward AI-native products, the compnay said.

Raja acknowledged that AI will definitely lead to some job losses, with certain roles being automated. However, he suggested that people who will succeed in this changing world are those who can effectively use these tools. In other words, the winners will be individuals who know how to command, manage, and work with AI systems rather than being replaced by them.