Suggestions
Tushar Singh
Exploring the intersections of technologies.
Tushar Singh is the Director of Research and Development at Freightcom, a position he has held since May 2023.2 Based in Waterloo, Canada, Tushar is an experienced professional with over 500 connections on LinkedIn.1
Professional Background
Prior to joining Freightcom, Tushar worked at Minute School, though his specific role there is not mentioned in the available information.2 At Freightcom, he leads the Research and Development team, which plays a crucial role in the company's operations.3
Recent Achievements
In December 2023, Tushar shared insights about Freightcom's accomplishments for the year. He highlighted that his team made a significant impact by building a platform that supported all of the products launched by the company during that period.4 This suggests that Tushar and his team are instrumental in driving innovation and product development at Freightcom.
LinkedIn Presence
Tushar maintains an active professional presence on LinkedIn, where his username is "therealtushar".14 This platform likely serves as a means for him to share industry insights, connect with other professionals, and showcase Freightcom's achievements.
It's worth noting that while Tushar Singh is a common name, this particular Tushar Singh should not be confused with others who share the same name, such as the Senior Research Analyst at the International Food Policy Research Institute (IFPRI).5
Highlights
Two years ago my team and I engaged with a client to assist in building a modern and scalable platform. The challenge was significant and one that my team and I relished as we had the opportunity to help shape the product daily. Over time our teams meshed and developed working relationships that come only through deep engagement and trust when solving difficult problems. As we neared the launch date so too came the realization that we would be looking for new challenges.
Then our client surprised me, they shared their longer-term roadmap with me and asked for my thoughts. Then they asked if we wanted to join them so we could build out as an internal team with greater resources and the strength of the organization behind us.
I am excited to announce that I have joined as Director of Research and Development at Freightcom. Joining me is my development team split between Waterloo and the GTA as well as several members of the current Freightcom development team. The impact that my team and I can make not only within Freightcom but with Freightcom on a global scale are opportunities that do not arise every day and excite me.
A significant component in my decision was the opportunity to continue collaborating with Turab (Tony) K., Jabir Alidina, Prasanna ('Priya') Thallamudi, CSPO®, CSM®, Michael Rochon, Irfan Kermally, Malia K., and the entire Freightcom team. Thank you to them for welcoming me and my team. I am truly excited about what we are building together and look forward to sharing the journey with you.
This post was NOT written by ChatGPT.
We're on the cusp of an #ai and #ml gold rush. Not only in the sense of large companies providing solutions but of mass market usage and product development. This has become possible as many processes have had digital transformations. Microsoft has taken a clear lead in this space and it's unlikely that their methods of integration into the Office Suite will be usurped by others. My mind turns to the gaps of where technology innovation will turn to next.
I believe the next major area of advancement will be AI@Edge. What this means is a highly available, even offline, systems that are operating on the edge devices. We have seen just how poorly Alexa, Siri, and Google Assistant perform with delays in responses as they connect to powerful cloud servers to process our requests. There have been significant challenges in making the assistants usable in realtime and requires models to run efficiently on device.
I'm not the only one thinking about this as a few weeks ago Meta's LLaMa model was leaked and has had hackers/tinkerers getting it running on a variety of different platforms (https://lnkd.in/ggKCw75Q). The move towards edge is key as it reduces the delay in data processing and also makes real time usage a reality for the everyday customer. It also opens up the possibility of data consumption and processing from edge devices and sensors without relying upon data center centric processing. Devices communicating more effectively in a mesh environment can act on a greater amount of sensor data to learn and act automatically without needing to share data to the cloud.
Companies like Apple recognize this and are actively hiring ML developers who are proficient not only in Python but also C/C++. A device centric company like them is looking to place Siri on device and also looking to make Siri generally usable. Google, Amazon, and others are doing the same. For both of these companies their efforts extend beyond simply the hand held devices. They have made major pushes into many "edge" industries such as automotive (CarPlay/Android Auto), home IoT (Matter), wearables (watches), and not to mention their extensive developments in fulfillment.
This pairing of edge devices with onboard models is the next major area of development. The miniaturization of models and getting them running on ever smaller devices will likely have a hardware component as well.
So while the gold rush in commercializing cloud connected AI has started the true leap forward will happen when we have AI@Edge. I have a feeling that it's not that far away and will transform how we live.
