I create content about deep learning | The DevRel and Deep Learning Dude

Hacking on LLMs, curating multimodal datasets, and writing a book about RAG.

Harpreet Sahota 🥑

i've literally been waiting for years to have an OPEN SOURCE model like qwen3-vl-embedding

i tested its multimodal retrieval capabilities on a corpus of 412 short video clips and the results literally blew my mind

here are the queries i tested:

1. a cartoon guy drinks merlot wine

i like this query because we see how it can retrieve based on semantics (a cartoon), text (the label merlot), and temporal action in the video (the cartoon guy drinks the wine mid-way through the video)

2. garfield runs out of the door

semantics are awesome here, the model knows i'm talking about the cartoon character...

3. a woman in a blue dress sleeping on a red bench

the skeptic in you might think that it's retrieving just based on the red color block in the 2/3 of the video...but notice the specific part of the query "a woman in a blue dress..." and this is only shown in 3s out of the full 10s

this is such a huge release 

and it's gonna open up SO much more for multimodal video retrieval this year

on my wishlist is natural language search of pcd datasets, who gonna ship that?

you can hack around with the model using the resources below

check out the docs: here: https://lnkd.in/gh7yfBNP

and the quickstart nbhere:  https://lnkd.in/gbinsx7p

🤖 Hacker-in-Residence @ Voxel51| 👨🏽‍💻 AI/ML Engineer | 👷🏽‍♀️ Technical Developer Advocate | Learn. Do. Write. Teach. Repeat.

for those who missed it, or were there in-person and know the trouble i was facing 😆, my ODSC webinar is now live on youtube!

in this session i discuss visual ai techniques for automotive damage detection. 

i used the cardd dataset, which is a public dataset built specifically for vehicle damage analysis and demonstrated how fiftyone helps you explore, visualize, and build models to detect dents, scratches, cracks, glass shatter, tire flats, and broken lamps.

i feel like its a super practical workshop with some insightful workflows for visual AI in automotive settings, but as the creator i might be biased.

only one way to find out for yourself, watch the webinar!

https://lnkd.in/ewfz5vZD

Open video: Mastering Visual AI with Vision-Language Models & Advanced Evaluation Techniques by Harpreet Sahota on youtube.com

Mastering Visual AI with Vision-Language Models & Advanced Evaluation Techniques by Harpreet Sahota

sometimes i look back at the shit i get done and i genuinely impress myself

how teams really build Visual AI in the wild? what's the he messy reality?  

i don't know...and maybe its because nobody’s asking practitioners and compiling their answers so we can find out the landscape actually looks like. 

so, we at Voxel51 are kickin off the...

wait for it...

 State of Visual AI survey...vey...vey

for those out ther building visual AI systems i want your hot takes

what’s the hardest part about making visual ai a reality? 

let us know in this survey

it'll take you 30 minutes and is a much better use of time than baby sitting your runs and staring at your loss curves

i can't offer much beyond my heartiest thanks, but you will get early access to the full report with benchmarks, common pain points, and where you stand within the landscape. 

here's a link: https://lnkd.in/gnsADACy

MedGemma 1.5 supports detection, but for best results, you'll need to fine-tune. 

there's a kaggle competition using the model, i created a starter notebook to give you a jump start on how to fine-tune it for detection

checkout the starter notebook and additonal resources 👇🏼

cc Omar Sanseviero Megan Risdal Logan Kilpatrick

in the asu area and wanna get started with open source contributions? then don't miss this hack

On March 21, 2026, we’re hosting Agents World: Visual AI Hackathon @Arizona State University, together with Pavan Turaga 🚀 

This is a hands-on day: meet people, build with AI, eat pizza, and leave with something you created after a short workshop and a full day of hacking. 
You’ll work with the FiftyOne ecosystem and real models used in practice, like Qwen, Gemini Vision, Sharp, and more. 

If you’re an ASU student who wants to build, learn by doing, and work with real Visual AI tools, this is for you.

Registration is open:
https://lnkd.in/gBCHW5Fg

Hope to see you at ASU. I’ll be there with the amazing Voxel51 team:  Sid Mehta, Ritchie Martori, Jimmy Guerrero ! 🙌

looking forward to the next meetup in boston!

Join our in-person meetup at the Microsoft Research Lab at the Massachusetts Institute of Technology in Boston, Massachusetts on Feb 18 to hear talks from experts on cutting-edge topics across AI, ML, and computer vision. Register here: https://lnkd.in/gi4kpSMB

Talks will include:

• SNAP: Towards Segmenting Anything in Any Point Cloud - Hanhui Wang at Northeastern University

• Culturally Adaptive AI - Anku R. at Massachusetts Institute of Technology

• Data Foundations for Vision-Language-Action Models - Harpreet Sahota 🥑 at Voxel51

• Better Together: Leveraging Unpaired Multimodal Data for Stronger Unimodal Models - Sharut G. at Massachusetts Institute of Technology

• Neural Radiance Fields for Image Verification - Sage Simhon at Pasteur Labs & ISI

#computervision #ai #artificialintelligence #machinevision #machinelearning #datascience

Syed Abdul Sami Rizvi, Matthew Ebisu and 118 other attendees

View event: Feb 18 - Boston AI, ML and Computer Vision Meetup

Feb 18 - Boston AI, ML and Computer Vision Meetup

Inspired by the amazing Hugging Face’s 2025 AI Timeline, we created one for the Fiftyone community. The FiftyOne Community Development Timeline highlights more than 131 contributions from 2025, including datasets, models, plugins, and integrations.

You can filter by type or modality and jump straight to each integration.
https://lnkd.in/ev2kpCGN

Big thanks to the entire Voxel51 community for building all of this together.
Harpreet Sahota 🥑, Daniel Gural, Paula Ramos, PhD, Jimmy Guerrero, Derek Van de Streek, Nicholas Lotz 🚀

Gonna be a great series of talks, lots of technical depth!

Chief Community Officer @ Voxel51 | Startup Advisor

Join us on Feb 18 for the Boston AI, ML and Computer Vision Meetup! Pre-registration is mandatory to clear building security - https://lnkd.in/gHtBuH49

Talks will include:

* SNAP: Towards Segmenting Anything in Any Point Cloud - Hanhui Wang at Northeastern University

* Culturally Adaptive AI - Anku R. at MIT

* Data Foundations for Vision-Language-Action Models - Harpreet Sahota 🥑 at Voxel51

* Better Together: Leveraging Unpaired Multimodal Data for Stronger Unimodal Models - Sharut G. at MIT

* Neural Radiance Fields for Image Verification - Sage Simhon at Pasteur Labs & ISI

***********

Want to build better computer vision models? FiftyOne is an open source toolkit from Voxel51 (our Meetup sponsor) that helps you curate datasets, evaluate model performance, visualize embeddings, catch annotation errors, and eliminate duplicate images—all in one place.

“pip install fiftyone” is all it takes to get started - https://docs.voxel51.com/

#computervision #ai #artificialintelligence #machinevision
#machinelearning #datascience #physicalai

NRG Energy

I work on the Innovation Optimization team, specifically focusing on product recommendations. For example, when someone signs up for electrical service, we can recommend non-commodity products, like surge protection or carbon offsets.

My role is building better systems to show people products they are more likely to be interested in, specifically focusing on the pipelines that combine eligibility and calculated purchase propensities to feed different recommender systems across channels and brands. It's fun! I'm also a data quality nerd - I like Great Expectations and Evidently AI.

Along with the items I mentioned above, some of the tools I use are: Python, SQL, Databricks, S3, Airflow, MLflow, and Azure DevOps.

Sr. Data Scientist

LendingTree

My two favorite accomplishments in this role:
- Teaching non-analytics folks how to use Tableau and our self-serve data marts. I did this through office hours and short tutorial videos. People asked good questions and were quickly able to apply what they learned. Very rewarding!

- Deploying a new expected lead value model, which allowed the team to more accurately assess value of outliers. It was a good collaboration across teams, and I was so glad to get it across the finish line and into production.

In general, I oversaw analytics for the Small Business vertical, including task prioritization and assignment.

Notes to self about other greatest hits: Match Check, new ELV calculation, automating EV process with clustering, SQL Sheriff assessment, Noob's Guide to ASH, ELR search engine campaign levels, Redshift eulogy with ChatGPT, Jon's son's storytime, Dynamic Routing w/ M&Ms

Business Analytics Manager

I have a lot of flexibility and autonomy in my role, so things change a lot from week to week, but here are some of my key responsibilities and accomplishments so far:

- Oversaw rollout of lead scoring model and application to dynamic routing, enabling the sales team to more effectively sell high quality leads. I continue to work with the sales team to ensure the model is performing well and helping them get the results they need.

- Implemented lead routing optimization tool, creating better opportunities for consumers and improving gross profit margin. I led the project, which required effort from two dev teams and the Product team. Besides project management, I also validated the data, quantified the business impact, and secured buy-in from the VP of Analytics, GM of Small Business, and Corporate Head of Sales before implementation. It was a great project, and I'm looking forward to building on its success!

- Working closely with our data engineering team, I led the migration of our lead revenue estimation model for three product verticals from legacy marketing tracking system and updated ETL process to incorporate finer grain lead quality attribution, enabling marketing channel managers to lean into higher quality traffic.

- I synthesize findings and provide actionable recommendations to leadership. Sometimes these are monthly marketing recommendations, and sometimes they are routing process improvements that will directly impact the bottom line.

- I recently used tree-based algorithms and funnel analysis tools to identify key upstream indicators of down-funnel conversion. We can use these indicators to modify our marketing efforts, improve our user/product interaction, and improve traffic quality for business partners.

- I love streamlining processes. For a Q4 '21 OKR, a colleague and I reduced annual seasonality index calculation processing time from a 1-2 days to 1-2 hours using automation tools.

Sr. Data Analyst, Strategy & Analytics

Black Bamboo Data

• Developing machine learning models for market research classification and segmentation
• Interpret KPI and digital advertising performance data and provide business recommendations
• Wrangling, analyzing, and visualizing data with Python (pandas, sklearn, numpy, and scipy) and SAS 9.4
• Deployed classification web application using Python and Streamlit
• Created Salesforce report suite for quarterly metric reviews

Data Scientist

Eva Garland Consulting, LLC

Data Analysis & Process Improvement:
•	Automated over 95% of lead collection data entry, saving hundreds of hours per year
•	Addressed low metric visibility in the Sales team by creating manager dashboards in Salesforce to track Sales KPIs
•	Prepared and delivered industry SWOT and Competitor Analysis
•	Used data from multiple sources to prepare and deliver 2019 Annual Business Development Report to BD/Sales Manager and CEO
•	Implemented Google Analytics and used collected data to measure organic and paid search campaign success
•	Redesigned workflow management tool and saved 100 hours of work per year

Other Duties:
•	Led sales calls with prospective clients in biotechnology and healthcare industry
•	Provided training to team members on workflow management tools and sales/marketing processes
•	Built the "SBIR Land" website to serve as a resource for potential grant requestors
•	Attended tradeshows (eg BIO, RESI, AUTM) to build brand awareness, collect leads, and develop relationships with potential clients

Marketing & Sales Analyst

PostMaster Social Media

•	Created and managed social media ads, sales funnels, and websites for clients
•	Conducted A/B testing on ads, improving click-through rate 5x
•	Managed and updated WordPress websites for clients, including significant functional changes, such as adding e-commerce (WooCommerce) and migrating from Shopify to WordPress
•	Coached client on creation of inbound lead generation campaign content that grew to be a significant ongoing source of clients
•	Content cited by well-known blogs, including The Manifest and Fupping
•	Created consistent video content, analyzing and simplifying complex material into understandable terms for channel subscribers
•	Managed relationships, including client calls and email contact

Web Developer

O.C. Tanner

•	Facilitated Kaizen Blitz focused on defects in Receiving/Purchasing and identified $39,000 of waste. Blitz team implemented redesigned process over 90 days
•	Led micro-blitz to improve international on-time delivery reporting and communication for a multi-billion dollar client
•	Worked with Supply Chain executives, managers, and team members to control project scope and ensure timely implementation of improvements
•	Wrote and delivered monthly Supply Chain newsletter and continuous improvement videos to 1000 employees across the enterprise

Notes for myself: Teaching lessons with the tennis ball game and the manager with oven mitts and paper sorting game, changing strategy deployment presentation mtg (timer, phones, blocking seats prank), fun video celebration for Phil's 40th bday, newsletter stats and saving praise emails in a separate folder, leading data collection and summarizing it for casting kata

Kaizen Blitz Leader (Process Improvement Leader)

•	Managed 2 production teams of up to 20 people; responsible for strategy deployment project implementation
•	Managed all front-line data collection, visualization, and analysis of 6-month metal casting quality Hoshin Kanri project
•	Used gamification to drive KPI's of safety, improvement, and recognition
•	Developed and delivered training activities on the 8 Wastes, 1-Piece Flow, 5S, and Team Management that were rolled out to hundreds of employees across the enterprise
•	Hosted a continuous improvement video series showcasing company-wide kaizen activities

Operations Group Leader

Salt Lake County

•	Inspected food establishments for code compliance.  Coached core staff on required improvements.  Conducted follow-up inspections as needed
•	Completed FDA-mandated science coursework
•	Used SigmaXL, Excel, and Oracle Discoverer to process data and create a shareable route optimization tool, avoiding a $9,000 cost incurrence on a new route management system
•	Made an Excel tool to optimize routes and confirmed statistically significant improvement in efficiency

Inspector, Department of Health

Utah State University

•	Full program life cycle management of an intensive English immersion program
•	Proactively identified and implemented program improvement opportunities
•	Researched technology shifts in the industry and applied trends to enrich the program
•	Created new programs from the ground up, including hiring staff and vetting curriculum
•	Maintained relationships with internal and external customers, including international government officials and NGOs. Supervised part-time team members
•	Tracked program budgets of up to $700,000

Project Manager

•	Promoted to full-time staff in less than a year
•	Organized passports and prepared J-1 visa documentation for intensive English program participants
•	Standardized program application process, increasing accuracy and reducing turnover time
•	Updated and maintained program website and electronic communications
•	Supervised 3 other part-time summer workers
•	Inventoried and organized bedding, home goods, and school supplies for over 140 people

Project Coordinator

My personal brand can be summed up in one word: originality. I am at my best when coming up with creative ways to solve problems or sharing big ideas with others, combining creativity and my diverse experience to engage technical and non-technical audiences.

Ambiguity is not a problem for me. Where some may shrink from very open-ended problems, I embrace those challenges as opportunities to innovate. Recently, upon joining a two-time Inc. 5000 scientific consulting firm, I quickly recognized an opportunity to use my unique skill set to take the firm’s analytics to a new level. I combined my love for data and creative problem-solving to build the sales team’s analytics infrastructure, establishing monthly metrics, web analytics, CRM dashboards, and annual reporting. I also automated over 95% of client data entry, resulting in improved data reliability and more time for sales team members to focus on client needs.

I am fascinated by statistical process control and continuous improvement. I am certified in Lean Six Sigma and worked as a Kaizen Blitz Leader for one of the top Lean manufacturing companies in the U.S. During that time, I led a cross-functional team on a project to reduce losses due to receipt of damaged products, resulting in $39,000 in cost savings. Identifying the project and building the team improved my project management capabilities, and facilitating the group discussion developed my people management skills.

Outside of work and school, I am often found contributing to the data science community on LinkedIn, building my storytelling skills, playing laser tag with friends in virtual reality, and traveling the world with my wife.

taplio

The Battle

Harpreet Sahota 🥑

Eric Sims

Followers

Posts

Engagements

Likes

Performance ScoreHarpreet Sahota 🥑's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Harpreet Sahota 🥑

Eric Sims

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Harpreet Sahota 🥑's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreHarpreet Sahota 🥑's LinkedIn Performance