Founder journey
Pavan Kumar Athreyapurapu
Serial Technology Entrepreneur | Founder & CEO, Sastra Innovations | Enterprise Systems Architect
More than 24 years across software engineering, enterprise customer delivery, architecture, entrepreneurship and product development.
Begin the journey- 24+
- Years of professional experience
- 2008
- Entrepreneurial journey began
- 2012
- Sastra founded
- ~350
- Projects delivered since 2012
My career has moved across software engineering, enterprise account management, industrial automation, product architecture, global delivery and company building.
That range shaped how I approach technology today: as a complete business system connecting customer need, architecture, project economics, people, execution and long-term commercial viability.
Before entrepreneurship
Technical and commercial foundations
My early career spanned software development and enterprise account management at IBM. That combination exposed me to both the engineering side of technology and the commercial reality of understanding customers, managing accounts and connecting technical capability with business outcomes.
It later shaped my view of entrepreneurship: not as isolated product development, but as the complete process of identifying a problem, designing the system, estimating the economics, building the team and delivering a useful result.
Foundation lessons
Enterprise customer understanding
Commercial awareness
Technology-to-business translation
Entrepreneurial beginning · 2008
From engineering to entrepreneurship
My entrepreneurial journey began in 2008. One of my first initiatives involved bootstrapping an industrial automation product for a recycled-paper mill, an environment where software, machinery, operators, process discipline and business outcomes had to work together.
I later served as Co-Founder and CTO of ATExpert I Services, taking responsibility across technology strategy, architecture, engineering, customer requirements and business execution.
What changed
Complete outcomes
I moved from implementing requirements to taking responsibility for the full result.
Operational reality
Technology must work with people, equipment, workflows and business constraints.
Founder responsibility
Architecture, delivery, economics and customer expectations cannot be separated.
Sastra · Since 2012
Building Sastra since 2012
Since 2012, I have built Sastra as a founder-led technology business, developing products and delivering approximately 350 projects across multiple industries and international markets.
The work taught me to evaluate technology not only as a system to be built, but as something that must be useful, commercially viable, operationally practical, secure and maintainable.
Founder responsibilities
Customer discovery and enterprise sales
Solution design, estimation and pricing
Product strategy and architecture
Hands-on engineering and technical leadership
Project financial and operational oversight
Go-to-market execution
Talent and remote-team coordination
Delivery governance and client relationships
The product journey
Lessons across products
Alongside client services, I conceived, built and operated products across education, healthcare, media, recruitment, customer-flow management, conversational software and interactive learning. Each revealed something about markets, users, economics, workflow design and where further investment was justified.
Institutional Systems → Real-Time Workflows → Content and Audience → Recruitment Decisions → Grounded Conversation → Governed Enterprise Reasoning
Institutional software · 2014–2022
Market timing and focus
Institutional products
2014–2017 · Discontinued
iSkool
A school management system for academic administration and institutional operations.
- Institutional records and structured data
- Role-based access
- Recurring academic and administrative workflows
2016–2017 · Discontinued
iHMS
A hospital management system designed around administration and operational workflows.
- Departmental coordination
- Sensitive operational information
- Reliability and accountability
2016–2022 · Discontinued
SastraQ
A hospital token and queue-management platform for patient flow and service coordination.
- Real-time patient flow
- Status transitions and queue visibility
- Physical service delivery connected to digital workflows
The investment decision
The products recovered development and maintenance costs, but entering mature markets late meant meaningful differentiation required reinvestment without a sufficiently attractive return.
A product can be technically successful, operationally useful and commercially break-even while still lacking a compelling future investment case.
Digital media · 2017–2024
Traction is not sustainability
PaperDabba began as a digital news portal and later expanded into a YouTube news channel, building an audience of approximately 25,000 subscribers.
PaperDabba lessons
What worked
- Organic audience growth
- Multi-channel content operations
- Approximately 25,000 subscribers
What did not
Audience traction did not translate into a sufficiently sustainable production-cost model.
The decision
I paused PaperDabba at the end of 2024 to concentrate capital, engineering capacity and founder attention on Medha, SastraPDF and MedhaOS.
What carried forward
- Popularity is not profitability
- Traction does not guarantee healthy unit economics
- Strategic focus sometimes requires pausing a visible product
Recruitment technology · 2018–Future
A deeper recruitment problem
The original Jobzfi platform operated from 2018 to 2022. It did not reach commercial break-even, but it exposed a deeper problem in recruitment technology.
Jobzfi lessons
The limitation
Conventional systems can reward résumé formatting over capability, miss transferable skills and misread non-linear careers.
The deeper problem
Recruitment technology had become efficient at filtering applications, but not necessarily effective at understanding human capability.
Connection to Medha
Those observations informed a capability-based reasoning model inside Medha.
The next generation
Planned relaunchJobzfi is intended to return with contextual evaluation, transferable-skill recognition and explainable recommendations.
Sometimes the right decision is not to abandon a product, but to pause it until the underlying technology can support the larger vision.
Grounded conversational systems · 2020–2022
Fluency is not enough
Flashbot was a first-response chatbot that answered questions only from content available on the connected website.
Flashbot lessons
Grounded responses
Natural-language access was useful only when responses remained tied to approved information.
Controlled scope
Useful AI needs clear knowledge boundaries and defined operating limits.
What it shaped
The experience influenced retrieval, source-aware reasoning and governed enterprise knowledge.
Purpose-driven product · 2021–2023
Purpose before permanence
Quizerrr was developed during the COVID period as a learning-from-home experience using quizzes, contests, battles, coins and rewards.
Quizerrr lessons
Purpose over category
It was a response to a specific social need, not an attempt to establish a permanent EdTech business.
Participation
Gamification and feedback can improve participation for children and adults.
A different measure
Success should be measured against the original objective, not only against permanence.
The learning curve
Lessons that shaped Medha
Technology alone does not determine whether a product succeeds. Market timing, customer need, distribution, user behaviour, operating economics, production capacity and strategic focus all matter.
Stopping, pausing or changing direction does not automatically mean failure.
Records and Roles → Institutional Workflows → Real-Time Operations → Grounded Conversation → Governed Reasoning and Execution
The current chapter
Building Sastra Innovations
Today, I apply this accumulated experience as Founder and CEO and as the hands-on enterprise systems architect behind Sastra’s current technology direction.
Current product hierarchy
01 · Persistent reasoning engine
Medha
Sastra’s proprietary reasoning engine for persistent context, multi-model reasoning, retrieval and governed tool execution.
02 · First commercial application
SastraPDF
The first commercial application of Medha, applying document intelligence across more than 70 document operations.
03 · Governance and control
MedhaOS
The control layer for identity, permissions, policies, approvals, auditability and private or hybrid deployment.
How I think
AI for real organisations
I focus on practical AI adoption inside real business workflows, not intelligence added as a cosmetic layer over disconnected processes.
Enterprise AI must operate within identity, permissions, tenant boundaries, policies, approvals, sensitive-data controls, auditability and human accountability.
The systems I build today are the accumulated result of more than two decades in technology and years of learning when to build, reinvest, pause and when a deeper problem deserves a new architecture.
Selective technology engagements
From idea to execution
Custom applications and internal platformsWorkflow automation and approvalsAPI integration and backend systemsApplication modernisationDocument intelligenceAI-enabled enterprise software
Writing and research
Ideas from experience
Future articles will explore enterprise AI, architecture, governance, product strategy, Decision Debt and lessons from building technology businesses.
Explore the Sastra blog