The Engineer Who Stayed in the Room
Most people who spend twenty years in enterprise technology eventually drift toward the boardroom - suits, strategy slides, market reports. Anandan Chinnalagu went the other direction. As CEO of MindPro Technologies, he keeps his hands on the technical architecture of the products his team builds. His PhD isn't decorative; it's operational. His academic research in machine learning directly informs CanString, TimeBound, and the AI integrations MindPro sells to clients across education, healthcare, and manufacturing.
The company he leads is headquartered in Karur, a mid-sized city in the Cauvery River delta of Tamil Nadu - not Bangalore, not Chennai, not the obvious place to build an enterprise AI firm. The choice is deliberate. Anandan has research affiliations at Government Arts College, Bharathidasan University, also in Karur. His corporate headquarters and his academic home occupy the same small city. There's a coherence to that geography that reflects how he thinks: local depth, global reach.
Building and sustaining MindPro as the top quality brand in IT and software development through continuous skills upgrading, innovative project delivery, and client-centric support.
- MindPro Technologies Mission StatementFrom Franklin Templeton to Founding His Own
Before Anandan built companies, he consulted for them. Between 2005 and 2008, he worked inside the enterprise software systems at Franklin Templeton Investments and Intuit - two firms where data architecture isn't a theoretical exercise but a business-critical infrastructure problem. Those years inside large financial and software organizations gave him a particular fluency: he understood not just how systems were built, but why they failed at scale, and what enterprise clients actually needed versus what vendors typically sold them.
He had already founded AC INFOTECH INC in Santa Clara in 2003, establishing a foothold in Silicon Valley's IT services market before most of his peers had figured out what "offshore delivery" meant. That company is still running. More than two decades after its founding, AC INFOTECH remains operational - a testament to Anandan's preference for building things that last rather than things that look impressive at a pitch meeting.
MindPro Technologies followed in 2008 - the same year the global financial crisis was busy dismantling the assumptions of the previous decade. Where others pulled back, Anandan expanded, establishing a full-lifecycle software development firm that would grow from a boutique shop in Karur into a 51-200 person operation serving clients in healthcare, education, manufacturing, and financial services.
AC INFOTECH INC Founded
Santa Clara, California. Twenty-plus years before "AI entrepreneur" became a LinkedIn archetype, Anandan was building enterprise IT infrastructure for Silicon Valley clients. The company is still running.
What MindPro Actually Builds
MindPro's service stack is deliberately broad but technically deep. The company works across five main domains: AI/ML/NLP integration using Google Cloud APIs (Vision, Speech, Translate, Natural Language), IoT consulting and managed services, mobile app development for iOS and Android, full web design and development including enterprise portals and e-commerce platforms, and business intelligence reporting using Pentaho, Jaspersoft, and BIRT.
The product portfolio tells a more specific story about where Anandan sees the future. CanString is a multi-channel notification platform with AI-powered sentiment analysis built in. CanString Class adapts that intelligence for educational collaboration. TimeBound is a pre-configured business intelligence platform. I-BIZ Suite handles enterprise workforce acquisition. These aren't consulting services dressed up as products - they're the result of Anandan applying his academic research directly to commercial problems.
CanString Analyzer
AI-powered content quality and sentiment analysis. Built on the same ML architecture Anandan published research on.
CanString Class
AI-driven educational collaboration platform. Notification intelligence applied to learning environments.
TimeBound
Business intelligence with pre-configured solutions. Data visualization without the six-month implementation lag.
I-BIZ Suite
Enterprise workforce acquisition software. Automates the recruiting and hiring workflow for mid-to-large organizations.
CanString
Multi-channel notification system spanning email, SMS, and push. The notification layer that knows how the recipient feels.
Expenses Log
Business application suite for enterprise expense management and workflow automation.
The Research Behind the Products
In 2021, Anandan co-authored three peer-reviewed research publications that placed him in a rare category: tech CEOs who don't just talk about machine learning but actively contribute to its scientific literature. Working alongside Ashok Kumar Durairaj at Government Arts College, Bharathidasan University, he tested sentiment analysis models against a dataset of 778,631 records drawn from Twitter, IMDB, Amazon, and Yelp.
The results were precise. Their custom fastText approach achieved 90.71% accuracy - outperforming a Linear Support Vector Machine (90.11%) and their SA-BLSTM model (77%) - while training significantly faster. The margin looks small until you consider the dataset size: the difference between 90.11% and 90.71% accuracy across 778,000+ records is not a rounding error. It's thousands of correctly classified sentiments.
Co-authored with Ashok Kumar Durairaj. Tested fastText, LSVM, and SA-BLSTM models on 778,631 datapoints from Twitter, IMDB, Amazon, and Yelp. FastText emerged as the fastest and most accurate approach.
90.71% Accuracy · 778,631 DatapointsCompared fine-tuned BERT against LSVM, fastText, BiLSTM, and hybrid models for sentiment classification across Twitter, IMDB, Yelp, and Amazon review datasets.
BERT Fine-tuning · Multi-dataset ValidationApplied SA-BLSTM sequence processing models to analyze sentiment and emotional patterns in social media conversations during the COVID-19 pandemic.
COVID-19 Social Media · SA-BLSTM ModelsThe Karur-to-California Arc
Anandan Chinnalagu holds a Master of Science in Computer Electronics and Applications and a PhD in Computer Science from Bharathidasan University in Tiruchirappalli, Tamil Nadu. That's the same university system affiliated with Government Arts College in Kulithalai, Karur - where he conducted his recent research. The degrees aren't just credentials; they're the connective tissue between his academic work and his corporate work.
His personal base is Sunnyvale, California - the heart of Silicon Valley's semiconductor belt, a few miles from Intel's original campus and AMD's headquarters. His company's engineering operations are in Karur, Tamil Nadu. He moves between these two worlds not as a commuter but as someone who built infrastructure in both places intentionally. The H-1B visa program, which he has spoken about publicly on NewsX, is not an abstract policy issue for him - it's the mechanism that makes his kind of transnational company building possible.
Since March 2017, he has also served as Chief Solutions Architect at CanString AI Products, operated through CrowdAround Inc. The CanString platform - a multi-channel notification and AI analytics system - is in many ways the commercial application of the research he would later publish. The academic papers and the product roadmap are pointing at the same problem: how do you teach a system to understand not just what a customer said, but what they meant and how they felt?
What He's Building Next
MindPro's service portfolio has expanded steadily into generative AI integration. The company's technical stack now includes Google Cloud AI APIs across Vision, Speech, Translate, and Natural Language - plus data analytics, forecasting, and connected device insights through their IoT practice. Their client verticals span healthcare, education, manufacturing, hospitality, and financial services.
The trajectory is toward tighter integration between the CanString intelligence layer and enterprise workflows. Sentiment analysis becomes workflow automation when you close the loop: a customer's negative review triggers a support ticket, which triggers an escalation, which triggers a resolution protocol. The academic research Anandan published in 2021 is the foundation layer of that architecture.