Interaction or Conversational Analytics is taking over how organisations understand and analyse customer conversations in an omnichannel environment. Manoj Kanumuri, (CEO, Sayint and Zen3) leveraged his vast experience in delivering AI, ML and digital user journeys for some of the world’s largest brands to build and scale the Sayint conversational AI platform.
Having spent most of his professional life in the US, Kanumuri got an opportunity to work on digital transformation journeys with many large corporations who were slowly adopting new tech initiatives. He noticed that most companies had large contact centres which were profit centres, but they had little understanding of what was happening in them. This is when it struck him that if he could put together a platform that can contextually understand contact centre conversations and provide stakeholders with a comprehensive analysis of all parameters they’d want to track, it would be the only missing piece in the customer journey puzzle.
We caught up with him to learn more about his plans to drive Sayint to greater heights with his team, currently consisting of 30 people.
What is Sayint?
Manoj applied his experience of leading a team of AI/ML researchers and Data Scientists at Microsoft to put together an incredible team for Sayint. The startup is almost 3 years old and has built capabilities to derive smarter insights from customer conversations across channels and languages.
We help organisations deliver better customer experiences by working inward and outward to understand, analyze and help relevant stakeholders take action on 100% of customer conversations,” says Kanumuri.
Sayint Aims to Be the Foremost SaaS Platform for Conversational Analytics
Headquartered in Seattle, and Hyderabad, Sayint is working with customers and partners in various industries to help improve, automate or change processes. Many clients like Microsoft, Teletext Holidays, TeamLease, Lutz.us and Ocean Holidays have seen incredible results with impact on revenues, customer experience, employee engagement and audit and compliance.
Sayint essentially functions on 3 levels. The first is to gather 100% of customer conversations from multiple sources including web, social, email, chat and contact centre.
The second step involves the automated speech recognition of the audio data and then processing analytics on all data.
The third step is how these actionable insights, derived straight out of conversations can be delivered which would create a positive impact for the organization. The delivery mechanisms include custom dashboards specific to stakeholders, voice bots, chatbots, email automation bots and RPA.
The startup has clients from BFSI, travel, HR and recruitment, and retail sector.
Conversational Analytics Is the Key
While conversational analytics is being adopted by a variety of enterprise clients in India, a few key region-specific challenges are yet to be solved because of the sheer number of languages and dialects in the country.
The move towards customer centricity and personalization is fuelling the push for the use of conversational analytics to understand the user journey, intent, sentiment, and actionable items to upsell, cross-sell or retain to customers and deliver better experiences.
This, Sayint believes is what they can enable with the platform.
How It Works?
Sayint has an incredible team of speech researchers, NLP engineers, linguists, data scientists and ML experts. They have successfully built an automated speech recognition engine on open source stack and are building ML models on top of it to support various languages, verticals and use cases.
Sayint also integrates with all major telephony, CRM and back end systems to support omnichannel analytics and easy relay of information.
In a traditional setup, companies typically do an audit of 2-3% of customer conversations. This random sampling is inefficient, and a lot of information is lost.
Sayint audits 100 percent of conversations and showcases the parameters in an objective manner to relevant stakeholders to take action, in real time.
Some of the use cases that Sayint solves as a platform, completely customized for each industry, vertical, use case or client are:
- Customer experience analytics
- Sales analytics
- Automated audit and compliance monitoring
- Automation using RPA and bots
- Web and Social Analytics
- Contact centre and agent performance analytics
Having served clients in multiple industries, Manoj shares some interesting use cases and their outcomes:
Human Resources (Staffing): Automated support requests and replies of over a million queries using Chatbot and Email Bot, automatically. This helped the client increase TAT from two hours to less than one minute and saved $600,000 in workforce costs over the last 6 months.
Travel: Improved average CSAT for Europe’s largest holiday provider by 10%; reduced booking error fines to 0 and delivered a GBP 300,000 per year saving; automated the entire audit and compliance monitoring process including redaction of PII Compliance information.
BFSI: Solved varied use cases in BFSI.
- Automated quality audits of 100% of customer interaction data
- Automated the entire feedback call process by deploying a conversational voice bot resulting in savings of $25,000 1 month after implementation
- Created fraud detection and compliance dashboards to give real-time alerts of anomalies.
Currently bootstrapped, the startup is constantly investing in building language models for different Indic languages while consciously acquiring clients and key partnerships that would help it scale. It has witnessed exceptional growth in the past few years with clients in different industries and verticals. “The journey so far has been interesting. Going forward, we intend to shorten our sales cycle, invest in speech tech and scale business to the US and UK with key partnerships already in place,” says Kanumuri.
Roadmap for next year
Kanumuri shares that Sayint has plans of introducing speech models in regional languages. “We aim for Sayint to become a SaaS and self-service platform, widen the integration net and see what tools we can integrate with to gain easy entry into prospect accounts. We look to make the best visualization to hit the screens to show aggregated metrics of every customer interaction,” says Kanumuri on a concluding note.