Federico Rostagno/123RF
 

AI In Supply Chain Management Use Case #3: LevaData

  • 23 June 2019
  • Sam Mire

This interview is part of our new AI in Supply Chain Management series, where we interview the world's top thought leaders on the front lines of the intersections between AI and supply chain management.

In this interview, we speak with Rajesh Kalidindi, founder and CEO of LevaData, to understand how his company is using AI to transform supply chains, and what the future of supply chains holds.

Rajesh Kaladindi

1.  What’s the story behind LevaData? Why and how did you begin?

RK: For myself and my sourcing team, the challenge was dramatic. Facing surging competition, I needed to find significant savings in the current year to defend gross margins. The directive: find four more points—a fifty-percent increase—beyond what we already negotiated four months ago. I was ready to act.

But there were no IT resources available. They were deployed across critical corporate initiatives. There would be no additional budget. After all, gross margins were already under pressure. And there wasn’t time to prepare another major sourcing event. Now the team had to innovate.

The team did have serious brain power and a talented consulting partner. Agreeing to a unique gain-sharing approach, the two groups huddled together. Over two months, they worked six days a week for twelve hours a day. They created a new data-driven negotiation approach. It combined every bit of relevant internal data with fresh, external market intelligence. The outcome: a comprehensive negotiation strategy and playbook.

Then, over one week, they engaged their suppliers. For every offer received, the team countered with real data. The suppliers had subjective arguments. The team had hard numbers, all at their fingertips in the playbook. Every opportunity documented, every insight captured.

At the end of the process, our team delivered almost double the savings requested by the product unit GM. That meant sixty million dollars added to the bottom line. Three months later we repeated the process for another business line.

That’s when the wheels started turning. This approach was perfect for a cloud-based solution, usable in a repeatable way. I reached out to forty other procurement leaders at other organizations.

All said the same thing: “That’s a problem we must solve.”

Many told me: “If you make it, I’ll buy it.”

So LevaData was born. The Cognitive Sourcing Platform offers customers these data-driven negotiating tools with the same focus on delivering a sustainable value proposition.

Our SaaS platform is purpose-built for the procurement and sourcing functions. It’s always on, which means it can find opportunities well before you need them. It uses crowd-sourced intelligence, unlocking the wisdom of the market for you. An innovative, AI-powered platform provides recommendations on when to act and how to optimize your negotiations. All this transforms sourcing & procurement into a sustainable competitive advantage with real, hard-dollar ROI.

Since 2014, we have brought a number of Fortune 500 companies on board (e.g., Bose, HPE, Nvidia, Lenovo, Fitbit, and Poly). In addition to helping procurement teams, we are driving the transformation of new product introduction processes by bridging sourcing intelligence with risk and cost optimization at the design stage. In fact, just last month, the company launched its New Product Information solution an extension to the company’s flagship Cognitive Sourcing Platform.

I have focused on helping these Fortune 500 companies with especially complex sourcing and manufacturing operations by incorporating LevaData’s AI-enabled tools into their procurement and NPI (National Provider Identifier Standard) processes to develop supply-sensing capabilities. In addition to helping these companies improve transparency and speed by providing a “single source of truth,” the platform’s predictive features quickly allow teams to identify emerging risks and procurement opportunities that provide sustainable cost savings.

Among these initial customers, I and my team have enabled a digital transformation within strategic sourcing.  All customers have achieved incremental annual cost savings of ten to 30 percent in the first year of adoption, with sustainable cost and risk management over time.  These companies are also reacting to market changes much more quickly, with agile response times dropping from three to four months to two to three weeks.

Instead of fighting fires, sourcing professionals using this cognitive platform can now shift from being reactive to proactive, leading tradeoff analysis of emerging risks and opportunities across supply chain, finance, and engineering teams.  These capabilities have been critically important in the past year, with fast-moving shifts in trade policy, tariffs and price volatility in major electronics commodities.

2. Please describe your use case and how LevaData uses artificial intelligence:

RK: LevaData applies a range of machine learning and deep learning to build more robust data sets, and pull insights and predictions from the data.

Examples include:

Data Cleansing: Item Classifier & Entry Deduplication

All AI-based systems require a large volume of comprehensive data that’s accurately labeled and regularly refreshed. Because completeness and accuracy are rare within an enterprise company’s procurement data however, LevaData’s item classification is a key early step in the partnership.

LevaData applies clustering approaches (such as K-means) to a company’s incomplete records of parts to identify each part and classify it by sub-commodity. Because there’s no universally accepted “master” classification system, LevaData has developed its own canonical information model and taxonomy hierarchy based on millions of part records pulled from our customers.

Predictive Costing

LevaData is the first platform to predict what the costs will be for specific parts (or a class of similar parts) between one to four quarters in the future. This model incorporates actual prices paid for a given part throughout LevaData’s network, third party “benchmarks” for part costs and a range of external market factors that may have impacted the price. The system then applies likely market factors to future quarters to estimate a narrow range of costs for that part. The most relevant benefit of this feature is providing recommendations on when to delay a purchase and when to buy in bulk.

Negotiation Savings Probability

This calculation, derived from machine learning algorithms, improves negotiation outcomes by giving the user a real-time estimate of their likelihood to reach a targeted cost. Procurement professionals typically use this probability in between rounds of negotiations with suppliers to understand how likely they’ll be to meet their target savings if they continue to push for a lower cost.

Risk & Opportunity Recommendations

LevaData continuously monitors multiple factors (marketplace activity, raw material input cost changes, news, risk events, tariff announcements, etc.) that have the potential to influence supplier costs, and sends alerts that are likely to be relevant to the sourcing professional.  Using neural networks and cluster analysis, new relationships are discovered across these dimensions on a continuous basis to further improve overall monitoring of emerging risks and opportunities in their supply base.  While the alert raises awareness of a risk or opportunity, LevaData also gives the user recommendations (using a combination of prescriptive analytics and business rules) on specific actions to take (such as identifying additional suppliers for a part that may experience shortages in the next few months).

3. Could you share a specific customer/user that benefits from what you offer? What has your service done for them?

RK: As Fitbit transitioned from a late-stage startup to a public company, their procurement team had to deal with volume-related challenges as well as the addition of more complex products. Moving from a manufacturer of discrete products to a health-tech platform resulted in exponential growth in the number of suppliers needed. To build a scalable team, Fitbit reworked its processes and brought on LevaData to reduce time spent on analyses and drive savings by bringing the “long tail” of spend under management.

To identify risks and opportunities, LevaData monitors Fitbit’s activity, transactions made by similar high-tech companies on LevaData’s network and third-party marketplace info. When Fitbit’s team chooses to take action by entering a negotiation, LevaData applies machine learning algorithms to generate “negotiation playbooks” customized to the required part and supplier.

Adopting a “single source of truth” for procurement spending and moving from Excel-based to automated, AI-driven analyses cut time spent on information gathering and analytics by an estimated 75 percent.

Relying on partially or fully-automated processes also allowed the sourcing team to move from semi-regular sourcing events (e.g., one to three per year) to quarterly events, bringing the percent of spend actively managed from 60 percent to 70 percent. As time spent on negotiation prep continued to decrease, the team introduced additional negotiations on an as-needed basis, increasing their spend under control to 90 percent.

Finally, the increased process efficiency, agility, and spend under management has translated into consistent incremental cost savings and improved negotiation outcomes.  Since 2017, Fitbit has identified and capitalized on over $30 million in opportunities identified by Leva, and realized an incremental two to three percent incremental cost savings on 90 percent of their direct material spend portfolio.

As they move forward in their procurement transformation journey, they are exploring how to extend the same applied AI insights in sourcing to their New Product Introduction (NPI) process to further optimize risk and cost at the product design phase. The expected benefits include faster overall time to market with new products, reducing risks and costs across the product life cycle, and improving competitive performance in the broader wearables and health solutions market.

About Sam Mire

Sam is a Market Research Analyst at Disruptor Daily. He's a trained journalist with experience in the field of disruptive technology. He’s versed in the impact that blockchain technology is having on industries of today, from healthcare to cannabis. He’s written extensively on the individuals and companies shaping the future of tech, working directly with many of them to advance their vision. Sam is known for writing work that brings value to industry professionals and the generally curious – as well as an occasional smile to the face.

Comments

COMMUNITY