HARMAN, a Connecticut-based Samsung subsidiary, specialising in audio electronics, launched HARMAN ForecastGPT, a predictive analytics platform that enables organisations to become more efficient through future forecasting and optimum resource allocation. It is aimed at understanding complex data patterns, and some of its chief features are advanced AI capabilities, real-time adaptability and the ability to seamlessly integrate with various platforms. It is compatible with any data format and source including CSV, Excel, SQL, and API.

HARMAN ForecastGPT’s Versatility 

It can benefit businesses in a plethora of ways including sales forecasting, supply chain forecasting, financial planning, and marketing even. 

“Embracing AI is imperative for business success and at HARMAN DTS, we are pioneering the application of AI to deliver tangible, bottom-line results. By understanding the unique challenges and aspirations of each client, we’re crafting AI solutions that go beyond generic predictions. The ForecastGPT platform is a testament to our commitment to equipping businesses with the tools to move past challenging roadblocks and fully capitalise on the potential of AI,” said Nick Parrotta, President – Digital Transformation Solutions & Chief Digital and Information officer at HARMAN.  

HARMAN & its Growing Capabilities 

HARMAN has a wide portfolio including audio and video systems, car audio, connected car solutions, professional audio and lighting equipment, and more. Additionally, it is also the parent company to brands like JBL, Harman Kardon, AKG, Mark Levinson, and Infinity Systems. 

“HARMAN’s data science team has made significant contributions by incorporating machine learning and deep learning models into a range of applications, such as predictive analytics, computer vision, NLP, and graph analytics,” said Dr Jai Ganesh, chief product officer of HARMAN, in an exclusive interview with AIM previously. 

The data science team of HARMAN employs both open source as well as commercial tools such as Python, TensorFlow, PyTorch, AWS, Azure, or Google Cloud Platform, Java, C++, Git, Jenkins, Docker, Kubernetes, R, Jupyter, SAS, MongoDB, Spark, Kafka, MySQL, RStudio, KNIME, RapidMiner, H2O etc.

The post HARMAN Introduces ForecastGPT, a GenAI Platform for Enterprises appeared first on AIM.

Connecticut-based Samsung subsidiary, HARMAN, is a global leader in audio, automotive, and connected technologies, offering a range of innovative and high-quality products and solutions to customers around the world. Founded in 1980 by Sidney Harman and Bernard Kardon, the company has grown to become a prominent player in the industry, with a strong reputation for delivering exceptional sound experiences, intelligent automation solutions, and seamless connectivity across devices and platforms.

Read more: Data Science Hiring Process at NoBroker

Harman’s portfolio includes a wide range of products and services, including audio and video systems, car audio, connected car solutions, professional audio and lighting equipment, and more. Harman is the parent company to a variety of brands like JBL, Harman Kardon, AKG, Mark Levinson, and Infinity Systems.

HARMAN empowers global enterprises with cutting-edge solutions that harness the immense potential of cloud computing, applied AI, data, IoT, advanced analytics, metaverse, quantum computing, cybersecurity, and 5G. The team uses data science to address a wide range of business issues in sectors like healthcare, communications, industry, retail, software, and hospitality.

Analytics India Magazine got in touch with Dr Jai Ganesh, chief product officer of HARMAN to know how the company implements AI in their daily lives and also their hiring process for talent and work culture. He is an alumnus of IIM-Bangalore and the University of Oxford. “We believe technology has the power to transform the world for the better and we build solutions that address some of the most pressing challenges facing enterprises and society,” said Ganesh.

Inside HARMAN’s Data Science Team

The AI, data and analytics team at HARMAN is 2500-strong. There is a centralised data science team as part of the chief product officer’s organisation, which is responsible for building AI-ML accelerators such as MLOps framework, building AI-ML based features and functionalities in products as well as creating proof of value demos. Each of their six key verticals consisting of healthcare, communications, industrial, retail, software, and hospitality has its own data science teams who work on client-facing projects.  

“HARMAN’s data science team has made significant contributions by incorporating machine learning and deep learning models into a range of applications, such as predictive analytics, computer vision, NLP, and graph analytics,” said Ganesh. These cutting-edge models enable HARMAN to enhance the customer experience and engagement by providing AI/ML-driven insights gleaned from various data sources. 

The data science team of HARMAN leverages both open source as well as commercial tools, applications and frameworks such as Python, TensorFlow, PyTorch, AWS, Azure, or Google Cloud Platform, Java, C++, Git, Jenkins, Docker, Kubernetes, R, Jupyter, SAS, MongoDB, Spark, Kafka, MySQL, RStudio,  KNIME, RapidMiner, H2O etc.

These models have diverse applications, ranging from predicting hospital readmissions with an accuracy rate of 93%, using over 50 inpatient data variables to identify risk factors, to powering conversational AI, recommendation engines, optimisation, and fraud-risk models. 

HARMAN has further cemented its position as an industry leader by unveiling its ‘Intelligent Healthcare Platform’ at CES 2023, which harnesses the power of AI and machine learning to provide actionable insights that improve customer engagement through predictive analytics. When it comes to customisation, HARMAN has also implemented AR and VR on JBL’s customisation page. 

Hiring Process

From freshers to senior positions, the company hires for different laterals. 

The interview process for hiring data science roles includes at least five rounds where the candidates are assessed on their conceptual, technical, problem-solving, team-playing, coding, and learning strength.

One of the most common mistakes candidates make is they don’t research well about the company before applying and focus only on technical skills.    

Expectations

HARMAN expects potential employees to have a strong foundation in programming languages such as Python, R, or Java. They should also be proficient in coding, debugging, and testing. It is critical to have a solid understanding of linear algebra, calculus, probability theory, and statistics to comprehend the underlying concepts of machine learning algorithms. 

Familiarity with ML algorithms such as supervised learning, unsupervised learning, reinforcement learning, and deep learning is essential. Experience in data cleaning, transformation, feature engineering, and normalisation is also important to prepare data for machine learning algorithms. Additionally, good communication skills, problem-solving skills, and a willingness to learn new concepts, algorithms, and technologies are required to excel in this constantly evolving field.

Read more:  Data Science Hiring Process at Park+

At the same time, employees can expect to have the freedom to think innovatively and work with a team of ambitious individuals from around the world and actively grasp the opportunities for learning, growth, and personal development. 

Work Culture

“HARMAN’s people are the biggest distinguishing factor that set the company apart from its competitors,” said Ganesh. 

HARMAN’s culture prioritises support, innovation, and excitement, and their diversity fosters innovative thinking. The company makes sure that you balance your personal and professional life well. Employees collaborate from different backgrounds to find innovative solutions and achieve technical successes. It offers employees a place to grow and feel like family.

Besides flexible working hours, hybrid office, and health insurance, HARMAN has other special perks for its employees, including the ReInventHers initiative, which focuses on aiding women who are resuming their careers after a break, and the AMIGO Maternity Engagement Program, which provides support to women employees during and after pregnancy. 

So, people who are comfortable with numbers who aim to make it big, maybe the right fit for HARMAN. 

Click here to apply. 

The post Data Science Hiring and Interview Process at HARMAN appeared first on AIM.

Back in October 2023, Samsung-owned HARMAN entered the generative AI race in the healthcare space with HealthGPT – a private LLM built on TII’s open-source model Falcon 7B. The latest version, HealthGPT Chat, is built on Llama 2. 

Healthcare is one of the most inherently complicated fields for generative AI to be harnessed given the vast amount of unstructured and sensitive data. 

“The primary motivation behind developing a private LLM like HealthGPT stems from enterprises’ concerns about data privacy and security. This issue arises particularly when using public LLMs, as they require transferring sensitive data to external entities, leading to potential uncertainties regarding how this data might be used,” Dr Jai Ganesh, chief product officer, told AIM, in an exclusive interaction. 

Ganesh further explained that another challenge with public models is their structural limitations, such as token restrictions and a lack of control over the entire processing chain. This lack of control can be detrimental, as any issues with public models can directly impact business performance. 

Understanding the severity of these issues, HARMAN entered the space of private models, coinciding with the open-sourcing moment of several foundational models around late February last year. It was around the same time when Meta’s LLaMA 1 was leaked online. Following this, models like Falcon 7B by TII and Llama 2 were released. However, the company chose to tap into the potential of open-source models instead of building an LLM in-house.

“We chose not to develop a foundational model ourselves due to the immense cost and resource investment required—estimated to be around $30 to $40 million. Instead, we focused on utilising existing open-source foundation models,” he commented. 

Dearth of Data

“We identified healthcare as the most relevant domain for our model due to the sector’s inefficiencies, the vast amount of unstructured data, and the challenges decision-makers face in deriving insights from this data,” said Ganesh. 

So given the scarcity of training data, the team found a rich source of such publicly reported data in clinical trial studies related to cancer, immune diseases, and heart diseases, leveraging this data to train HealthGPT. 

However, the problem is not confined to the amount of data available, but also its poor quality. Models trained on skewed datasets can lead to biased outcomes. “We have mechanisms in place to identify and correct bias in the data. This is not limited to the data preparation stage; the output is also scrutinised for bias,” he added. 

Rigorous testing frameworks, involving extensive querying, are integral to maintaining the integrity and privacy of processed data.

Tackling Hallucination Responsibly 

It is not new that LLMs are prone to spewing wrong information. But in critical areas like life science, where accuracy and reliability of diagnosis, such hallucinations can cost you serious threats. To overcome these risks, HARMAN uses a combination of automated mechanisms and human-in-loop intervention. 

HARMAN has a multifaceted approach to managing hallucinations. Firstly, HealthGPT employs guardrails to regulate the level of hallucination. “Initially, the accuracy of the models was around 74%, but with continuous refinement, it has improved significantly, reaching over 85-90%,” said Ganesh.

Secondly, the model’s interface provides control over settings like temperature and token number, allowing users to check the extent of hallucinatory outputs. Higher temperatures increase hallucination, while lower settings reduce it. Thirdly, human oversight is involved with the medical doctor of the team validating AI-generated results. This is supplemented by a feedback mechanism for a continuous loop of user input that helps in refining the model. Furthermore, RAG feature adds references to answers for better information credibility. Finally, the system includes a benchmarking section that compares the model’s performance with other studies and models. 

However, responsible AI is at the heart of HealthGPT’s success. “One of the reasons why we chose to go for private LLM is because it ensures end-user control. Unlike models hosted on unknown cloud instances, HealthGPT model operates within the user’s own Virtual Private Cloud (VPC) and cloud environment,” explained Ganesh, emphasising how the approach improves security and privacy as the model is fine-tuned on the user’s data within their controlled environment. 

Pre-fine-tuning checks are also implemented to detect anomalies, with a focus on privacy through automated mechanisms for handling Personally Identifiable Information (PII) and Protected Health Information (PHI). 

Harman’s approach in the healthcare domain, as well as others, is characterised by a human-centric philosophy. This approach involves understanding problems holistically and placing the human, or decision-maker, at the center of solutions. This philosophy is fundamental to Harman’s interactions with its global customer base, which varies from companies in the proof-of-concept (POC) stage to those in more advanced stages of implementation.

What Next

Although not yet deployed for live customers, HealthGPT has found compelling proof-of-concept (POC) stories showcasing its potential applications in diverse domains. One POC demonstrates the use of HealthGPT for personalised data analysis across sectors, allowing customisation for individual needs, such as drug discovery in pharmaceuticals. 

Another story involves using the model for extracting insights from medical instrument data, emphasising its capacity to handle large-scale, structured information. A third user story features a pharmaceutical company leveraging the model to bolster drug discovery by integrating clinical trial data and information from sources like PubMed. IQVIA, Roche, Aetrex are among the major clients that HARMAN serves.

“Currently, we are experimenting Mistral AI’s Mixtral 7B for future iterations. The aim is to constantly push the boundaries of auto foundation models,” said Ganesh. 

The company’s generative AI strategy involves integrating more diverse data sources with a focus on ensuring data quality and compliance with regulations like HIPAA. This necessitates extensive training for developers handling healthcare data. Alongside this, Ganesh and his team are working towards introducing multimodal features. 

But no matter how strong the generative AI product it comes up with, HARMAN’s strategy will always be characterised by a human-centric philosophy. This approach involves understanding problems holistically and placing the human, or decision-maker, at the center of solutions. 

Moreover, there are plans in place to expand HealthGPT beyond healthcare into domains like manufacturing and IT management, the other core focus areas of HARMAN, aligning with HARMAN’s strategic plans. 

The post How HARMAN is Solving Healthcare Problems with GenAI appeared first on AIM.