When geeks got out of the lab



Machine Learning

Artificial Intelligence

May, 2018


Discovering new business models around Machine Learning, Augmented Reality, Virtual Reality, and Internet of Things with technology teams at the NBA (New Business Accelerator)- Cyient’s internal startup incubator


Developed a clear business model around Machine Learning, and established a (quasi) process for startup discovery which can help other technology teams involved in similar projects


Learning Journey

Cyient is an NSE listed company, offering Engineering and IT services for 25+ years. With a vision to drive future growth with innovative solutions, they set up an entity called New Business Accelerator (NBA) which would function like an internal startup incubator. Having identified a few focus areas like Augmented Reality, Virtual Reality, Machine Learning, and Internet-of-Things, they set up teams with mixed technology skills. Now they were looking for a structured approach for discovering new business models, leveraging on methodologies like Design Thinking and Lean Startup.

TinkerLabs approach- give techies a flavour of startup life

From our initial conversations, we could gather that the teams come from a strong technical background, and have valuable experience of delivering on diverse projects. However, scope and deliverables of these projects were typically well defined. What they needed was an approach for identifying unmet needs and conceiving new solutions (products or services) to unlock business opportunities.

What happened on ground:

We designed a 3-month journey of learning by doing, and started with the team working in the machine learning area. The first touchpoint was a 2-day workshop, where the team was introduced to a) Design Thinking methodology for problem discovery and product/service development, b) Lean Startup concepts like smart experimentation, MVP, and pivoting, and c) business model canvas. This was done on a practice challenge area inspired from their common workplace experiences (around technology-enabled learning). The team engaged in empathy research with multiple stakeholders both inside and outside the organization, using methods like qualitative interviews, shadowing and user immersion. Based on their insights, they developed multiple solution concepts by using creativity techniques like constraint toggle and creative remix. Then, using approaches like storyboarding and wizard-of-oz, they prototyped their ideas and got user feedback. After tweaking their ideas to meet user expectations, they developed a complete business model for their chosen concept.

With this first-hand end to end experience of startup modelling, they prepared to apply the learnings on their live challenge- automation of feature identification in satellite images using machine learning. They first drew a broad ecosystem map with various stakeholders- existing and potential clients in the GIS space (like TomTom, Here, Google Maps), the GIS CoE (Centre of Excellence) at Cyient, and their business units focused on industries like road navigation, mining, oil and gas etc. They then set out for primary research to gather perspectives from these various stakeholders, with an intent to uncover unmet unarticulated needs/pain points. This activity spanned over 3 weeks. In the next touchpoint, we met for a 1-day workshop where we synthesized the various findings into 3 major opportunity zones, and defined certain experiments to validate the opportunities and some solution concepts. Again, over the next 4 weeks, the team developed some MVPs and tested them with intended users. During this period, they were tweaking their concepts after every user test.

In the final touchpoint, we narrowed down to two big opportunity spaces based on the user feedback, created a business model around them, and made an investment pitch to the management team. With this success story under their belt, NBA management is now onboarding other startup teams.

What did we achieve:

We were able to develop a clear business model and help Cyient validate their strategy of setting up an internal startup incubator.  We also established a (quasi) process for startup discovery which can help other technology teams involved in similar projects.

Highlight of the project:

We learnt as much about geospatial technology and machine learning from the Cyient NBA team, as they hopefully did about Design Thinking and Lean Startup! A lot of forgotten lessons about algorithms were refreshed, and it was a refreshing experience 😊

When Geeks Got Out of The Lab.

We unapologetically push boundaries for wow results