Sam Zheng, CEO and Co-Founding father of DeepHow, spearheads a quickly evolving startup, backed by esteemed buyers. DeepHow revolutionizes expert workforce coaching with an revolutionary, AI-powered, video-centric data capturing and switch platform.
Previous to DeepHow, Sam devoted over a decade to Siemens, driving digital innovation throughout numerous industries. His noteworthy initiatives, such because the Cloud Digital Inspection Jacket, have considerably improved technical data sharing, effectivity, and consumer expertise, incomes his group the distinguished Siemens Innovation Award.
Concurrently, Sam serves as an Adjunct Professor of Psychology at Tsinghua College and holds a Ph.D. in Engineering Psychology and a Grasp’s in Statistics from the College of Illinois at Urbana-Champaign.
You could have an academic background in psychology and statistics, how did you transition to specializing in movies and machine studying?
My background in psychology and statistics truly served as a pure segue into the realm of machine studying and video-centric platforms. The examine of psychology sparked my fascination with the human thoughts and intelligence, notably the method of abilities studying and experience growth. In the meantime, statistics offered the mathematical basis to discover synthetic neural networks, impressed by our organic mind.
In at this time’s digital age, movies have emerged as a extra participating, interactive, and efficient studying medium. This shift is clear with platforms like YouTube and TikTok, the place customers, notably the youthful technology, make investments hours consuming and studying from video content material.
Nonetheless, the standard course of of making educational or how-to movies, particularly the enhancing half, is time-consuming and labor-intensive. A brief video spanning a couple of minutes might require hours of meticulous work. Recognizing this inefficiency and the potential to reinforce the educational expertise, my co-founders and I made a decision to harness the facility of synthetic intelligence (AI) and machine studying to expedite the how-to video creation course of.
Our AI-powered video platform can remodel hours of laborious work into mere minutes, drastically enhancing effectivity with out compromising effectiveness. In essence, my tutorial grounding in understanding human cognition and the statistical fashions that mimic it helped pave the way in which for this revolutionary enterprise.
You could have a number of patents beneath your title, what’s a very powerful one that you simply labored on?
All my patents concentrate on harnessing expertise to reinforce human efficiency. Throughout my tenure at Siemens, one notable venture concerned engaged on an answer for Professor Stephen Hawking. We developed an intuitive eye-typing enter technique to help people with ALS, much like Professor Hawking’s situation. This revolutionary work is now a part of a pending patent.
Nonetheless, probably the most vital patent I’ve contributed to is a latest one: the Generative AI-Powered Know-How Administration Platform for Industrial and Manufacturing Organizations.
Here is a short overview:
Our invention presents a cutting-edge generative AI answer particularly tailor-made for industrial and manufacturing organizations. It successfully captures, organizes, and disseminates each institutional and tribal data, intersecting fields like synthetic intelligence, machine studying, coaching and growth, and data administration programs.
Industrial and manufacturing sectors typically face immense challenges in capturing, organizing, and sharing important data. Excessive employees turnover charges, complicated processes, and the perpetual want for upskilling amplify these difficulties. Conventional data administration strategies are sometimes cumbersome, time-intensive, and lack flexibility, necessitating a extra superior answer.
Our generative AI answer employs proprietary algorithms and machine studying methods to streamline the creation of video-based normal working procedures (SOPs), optimize workflows, and facilitate fast, environment friendly entry to data through AI-driven chat options. On account of its adaptability and scalability, our answer is suited to a variety of producing contexts.
Might you share the genesis story behind DeepHow?
Earlier than we launched DeepHow, I used to be working alongside our different founders, Patrik Matos da Silva and Wei-Liang Kao, at Siemens, driving a wide range of digital innovation initiatives in industrial and manufacturing sectors. Our journey took a giant leap once we joined the Techstars Mobility Accelerator in Detroit in 2018. The expertise was primarily a boot camp, serving to us validate our thought, determine gaps, and join with potential buyers, companions, and mentors.
We noticed the worth in abilities and experiences that individuals had cultivated over time however there was one problem- there wasn’t a really efficient approach to seize and share this experience. We noticed how quickly expertise was advancing, and it occurred to us that the strategies we used for worker coaching hadn’t saved up. We had been nonetheless counting on outdated, time-consuming strategies that weren’t environment friendly and, frankly, weren’t participating sufficient.
I spotted there was a singular alternative to mix advances in AI and video expertise to utterly change how we seize, construction, and share data. So we got down to construct DeepHow, a platform the place firms can create superb coaching movies in-house and stay in management. To actualize this concept, we created an AI system known as “Stephanie”. Stephanie is the spine of our answer, capturing specialists’ know-how and making a data repository. It generates step-by-step, interactive how-to movies to hurry up the educational course of for brand new and fewer skilled staff.
By the usage of AI workflow indexing and segmentation, we have managed to create video content material ten occasions sooner and enhance workforce efficiency by 25%.
We’ve come a good distance for the reason that early days, however our mission continues to be the identical. We’re dedicated to serving to firms leverage their wealth of institutional data and experience and allow their staff to repeatedly study and develop. It’s an exciting journey.
What are a few of the office challenges DeepHow is fixing?
Data Switch: In any office, transferring abilities and data, particularly from seasoned staff to new hires or much less skilled employees, might be fairly a problem. Now we have made it potential to seize and construction that data in a approach that is easy to share, damaged down into segments and simpler to know.
Coaching Effectivity: Conventional coaching might be tedious, proper? Lengthy hours, laborious to retain information- it is not all the time probably the most environment friendly course of. With DeepHow, staff can study at their very own velocity and in a approach that’s higher tailor-made to their very own distinctive studying type. It’s our aim to make coaching painless and pleasurable.
Abilities Hole: Typically, determining the place your groups’ abilities gaps are can really feel like looking for a needle in a haystack. That’s one other factor that we may help with. Our platform makes use of analytics to point out you the place there’s a hole in coaching and allows coaching groups to bridge that by creating the precise content material they want.
Outdated Coaching Supplies: Issues change quick, don’t they? Requirements and procedures, they’re all always evolving. Properly, we be sure that your coaching supplies are by no means left behind. Enhancing and updating is fast and painless, and with the ability to share these updates throughout totally different places and shirts with a easy abilities project makes certain that nobody is left with out probably the most up-to-date data.
On-Demand Studying: Comfort is king! Everybody desires to have entry to what they want, once they want it. That’s the philosophy behind DeepHow. We consider that coaching should not be a scheduled occasion that disrupts the complete day, however a versatile useful resource that’s there once you want it. Being confined by a spot and time isn’t preferrred. Have to study one thing particular, proper now? Search and watch. It couldn’t be simpler.
Boosting Worker Engagement: Everyone knows that coaching can generally be, effectively, not probably the most thrilling factor. But it surely completely doesn’t should be this fashion. DeepHow permits creators room to be artistic. Studying ought to be a enjoyable, participating and pleasurable expertise.
Fostering Communication: Communication might be tough, particularly when coping with a posh course of or process. Our platform makes communication easier by enabling step-by-step guides that enable staff to simply digest and perceive the duty at hand, selling clear and constant messaging throughout the board. Our platform additionally understands, interprets, and transcribes in almost 50 languages and counting. This characteristic alone has confirmed to be one of the vital invaluable instruments many firms possess. Permitting somebody to study of their native language ensures higher understanding and boosts morale.
How does DeepHow allow enterprises to create an adaptive coaching program?
Let’s think about the standard coaching atmosphere. You’ve obtained static materials, inflexible schedules, and a one-size-fits-all method. Now, these strategies don’t consider that everybody learns in a different way and at a special tempo. They don’t seem to be scalable or versatile to accommodate the quickly altering panorama or particular person worker progress. These are vital ache factors for any enterprise, proper?
That’s precisely the place DeepHow steps in. We enable you flip the script on these points. We allow companies to develop agile coaching packages which can be dynamic, personalised, and extremely attentive to enterprise wants and worker studying patterns.
Our platform captures knowledgeable data in easy-to-follow, video-based studying modules. However we don’t cease there. We harness the facility of AI to investigate staff’ interactions with these fashions, enabling deeper perception into the place your abilities gaps stay. It’s about turning weaknesses into strengths and capitalizing on particular person areas of experience to foster a steady studying tradition.
What function does DeepHow play in rising office security?
Security is such a important facet of any office, but it’s typically troublesome to get it proper, resulting in unlucky accidents and rules violations. That is sadly true in industries like manufacturing, development, or healthcare, the place even a small lapse can have vital penalties.
So the place does DeepHow match into this image? Properly, we’re obsessed with making certain that security practices are clearly understood and persistently applied throughout the board.
We do that by offering a platform the place firms can simply seize and share knowledgeable data on security protocols. As a substitute of the old-school, hard-to-follow manuals, we provide interactive, step-by-step video guides. They’re easy, straightforward to know, and most significantly, accessible anytime, wherever. This implies staff by no means have an excuse to not keep updated with the newest security protocols. Security ought to all the time be a high precedence and everybody deserves to really feel protected at work.
How is Generative AI used to allow this to occur?
On the coronary heart of our Generative AI-powered know-how administration platform, “Maven,” lies a set of revolutionary capabilities geared toward revolutionizing the way in which industrial and manufacturing organizations seize, handle, and share their institutional and tribal data. Maven employs proprietary algorithms and superior machine studying methods to simplify and improve a wide range of processes:
- Streamlined Video SOP Creation: Utilizing our distinctive Generative AI algorithms, Maven assists within the computerized technology of video capturing guides and narration scripts. This facilitates the creation of high-quality video SOPs with consistency, setting a brand new trade normal.
- Environment friendly Workflow Optimization: By leveraging superior machine studying methods, Maven’s workflow visualization instruments allow customers to simplify complicated duties, thus selling productiveness by way of user-friendly interfaces and good purposes of AI.
- Strategic Data Mapping: Maven’s AI algorithms assist organizations map essential know-how, determine data gaps, and uncover important content material alternatives. This leads to the event of focused coaching supplies uniquely tailor-made to the wants of a talented workforce.
- AI-Enhanced Chat Function: Powered by modern pure language processing algorithms, Maven’s AI Chat allows customers to entry essential data rapidly and effectively utilizing pure language queries.
- Multimodal Content material Era: In conditions the place a written SOP is unavailable, Maven can analyze video content material and generate multimodal content material like step-by-step video SOPs, textual content SOPs, and workflow diagrams, thereby streamlining the content material creation course of.
By marrying superior Generative AI with the newest data administration methods, Maven provides organizations a singular software to harness their expert workforce’s potential, driving steady enchancment and innovation.
Are there different sorts of machine studying algorithms which can be used?
Certainly, DeepHow leverages a spread of refined machine studying algorithms and AI methods inside the realms of Pure Language Processing (NLP) and Pc Imaginative and prescient. These methods, each supervised and unsupervised, underpin our proprietary, domain-specific AI expertise, which has been educated and optimized for the economic and manufacturing sector. Key areas of utility embody:
1) Workflow Segmentation: We make the most of machine studying algorithms to extract important data and steps from complicated, unstructured process demonstrations captured in movies. This allows us to interrupt down complicated procedures into manageable, teachable steps.
2) Multimodal Step Embedding: By modeling an ‘exercise genome’, we’re in a position to reconfigure instruction and workflow steering to higher swimsuit particular operational contexts.
3) Cross-Modality Retrieval: We use superior in-video search methods to facilitate multi-language, skills-based content material retrieval. This enables customers to entry related data with better effectivity and precision.
4) Know-How Mapping: We assemble a data graph that visually represents a company’s core competencies. This mapping permits firms to determine their data belongings clearly, enabling simpler upskilling and coaching methods.
These superior machine studying methods, mixed with our concentrate on industrial and manufacturing companies, enable us to supply a complete answer to the distinctive challenges confronted by organizations in these sectors.
For firms that want to get began, what’s the method?
We’ve designed our platform with simplicity in thoughts, so onboarding your organization doesn’t should be complicated. Actually, greater than 80+ enterprise and SMB prospects have efficiently deployed our options to over 400 websites in 24 nations throughout 6 continents.
First, our groups will meet and have a dialog about your organization’s particular wants and ache factors. We wish to perceive your objectives, the coaching challenges you’re going through, the sort of abilities your workforce wants – the entire image.
Subsequent, we information you thru the method of capturing your specialists’ data. This may be about any course of or ability that’s vital to your organization. Our group will help you in creating these step-by-step video guides utilizing DeepHow’s intuitive instruments.
Our group is there to assist you each step of the way in which, from the preliminary setup to the continuing optimization of your coaching program. We’re right here to kind a partnership with you and improve your workforce’s abilities and effectivity. Merely go to DeepHow.com to get began.
Is there anything that you simply want to share about DeepHow?
On the coronary heart of DeepHow is a transparent, compelling mission: we purpose to empower each expert employee to turn into an knowledgeable. We try to make data switch and coaching seamless, participating, and cost-effective, harnessing the transformative energy of AI. We firmly consider that expertise ought to increase human capabilities, not exchange them. This precept guides all the things we do.
In our quickly evolving technological panorama, this mission is extra related than ever. The shift in the direction of digital transformation and Business 4.0 has modernized manufacturing, introducing an array of superior applied sciences. These improvements have considerably altered job necessities, demanding staff to accumulate new technical abilities to function, preserve, and optimize these refined machines. The tempo of change is so fast that conventional coaching approaches wrestle to maintain up, resulting in an ever-widening abilities hole.
Our purpose is to deal with this problem head-on, empowering staff to ‘construct again higher’ by reskilling for the factories of tomorrow. Excessive ranges of automation imply there’s much less want for guide labor; as an alternative, the main target shifts to leveraging staff’ experience and instinct in working superior technological programs.
Factories have developed considerably over the previous decade, with the mixing of robotics, cobotics, and analytic applied sciences that frequently optimize output and decrease waste. To handle these applied sciences, an upskilled workforce is important.
DeepHow gives a contemporary coaching methodology, enabling producers to draw expertise, fill entry-level positions, and progressively upskill staff for superior roles throughout manufacturing, logistics, and planning. With an emphasis on up to date, participating coaching, we’re serving to to shift the notion of producing from a dead-end profession to a dynamic, technology-driven area with limitless potentialities.
Thanks for the nice interview, readers who want to study extra ought to go to DeepHow.