Publications

HOW CHANGES IN BOND CALL FEATURES AFFECT COUPON RATES. Journal of Applied Corporate Finance, 12: 92–99. doi: 10.1111/j.1745-6622.1999.tb00664.x 

Patents

• Cable fault detection - US 7725295 · Issued May 25, 2010

• Electrical substation monitoring and diagnostics - US 7558703 · Issued Jul 7, 2009

Projects

• Operating as the sponsor on a project for a Fortune 50 Oil and Gas customer looking to reduce unplanned outages and asset failures, I organized a team to develop an advanced analytics solution. ML models developed and deployed on client Azure tenant (alternatively as a SaaS) to help client’s engineers locate and analyze anomalies, contextualize the identified anomaly for potential failures on generation and compressor trains.

• Working closely with Microsoft's Manufacturing AI Innovation Council on industry use cases, I sponsored a team to create detailed use case definition, develop scope, architecture to couple machine learning anomaly detection with ChatGPT/GPT4, Large Learning Model (LLM) to generate custom embeddings and monetize Generative AI.

• To expand sales opportunities, I initiated a partnership with Snowflake and Databricks among others to create a win-win plan to double our qualified pipeline of opportunities.

• Leading an analytics project for a Japanese OEM looking to increase its asset/plant uptime, I led a team to conceptualize, develop advanced statistical and machine learning/AI model to predict remaining life of assets that impact uptime due to high failure risk.

• Led a project with a utility faced with potential bankruptcy due to aging assets causing wildfires. I built a team to define scope, architecture (Azure and GCP) and train a CNN AI model with high precision & recall. The project received Gold Cannes Lions Award at the Cannes Lions International Festival of Creativity for Creative Business Transformation - Creative Application of Emerging Technology.

• A major utility was looking to monetize their smart meter rollout through predicting meter health, identify roof top solar generation and identify electricity theft using smart meter data. I led a team to scope, build machine learning models on Azure to identify patterns consistent with the use cases.

• A major utility was facing a failed data lake & analytics platform project with a competitor. I was brought in to assess recovery of the project and then take it over. I led recovery assessment including developing cost, scope, schedule, architecture, resource and subsequent execution. The project built on-premise Hadoop, Kafka/NiFi based data lake to provide normalized structured, unstructured, time series and streaming data to analytics consumers. The value exceeded over $20 million and was executed as an Agile delivery.

•  A large utility was looking to optimize procurement of natural gas to meet daily generation demand. I led a team to develop and successfully deliver a solution that performed network optimization taking various procurement constraints in to optimize billions of dollars of natural gas purchase annually.

• To transition to a major analytics player, I built a team from scratch and delivered Asset analytics, T&D, Demand Analytics solutions to analyze asset health, asset risk, electricity theft, outage, vegetation and analysis of DER penetrated feeders based on machine learning, Bayesian and physics-based models.

• Operating as a technical lead for a Bay Area start up, I successfully ran the due diligence with VC’s to raise Series C and D funds of $46+ million.

• Hired to set up an Advanced Applications team at GE and implemented Agile development with TDD, CI/CD and Agile project management, I worked with product management and sold GE's executive management on an analytics vision to leap frog the competitors. I developed a plan to partner & develop analytics on C3 cloud platform which catalyzed GE's Predix platform.

• During the war in Iraq, the US Army Corp. awarded distribution SCADA project to ABB. As the sponsor and leader, my team designed communications network, manufactured, delivered and commissioned 75 substations while the war was still ongoing.

• As the R&D program manager, I worked with ABB's corporate research to develop and deliver a first-of-its-kind Neural Nets (AI) running on an embedded hardware to detect fallen, live electric wires (high impedance faults) Neural Net in an embedded device