Feed aggregator

BYDFi's Innovative Trading Approach

Hacker News - Fri, 04/11/2025 - 5:02am

Article URL: https://www.bydfi.com/en

Comments URL: https://news.ycombinator.com/item?id=43651847

Points: 1

# Comments: 1

Categories: Hacker News

The Met Police is set deploy permanent live facial recognition cameras on street furniture in Croydon from summer 2025, but local councillors say the decision – which has taken place with no community input – will further contribute the over-policing...

Computer Weekly Feed - Fri, 04/11/2025 - 4:44am
The Met Police is set deploy permanent live facial recognition cameras on street furniture in Croydon from summer 2025, but local councillors say the decision – which has taken place with no community input – will further contribute the over-policing of Black communities
Categories: Computer Weekly

JØKU – Balatro inspired game built with AI (mostly)

Hacker News - Fri, 04/11/2025 - 4:15am

Article URL: https://playjoku.com/

Comments URL: https://news.ycombinator.com/item?id=43651524

Points: 1

# Comments: 1

Categories: Hacker News

Show HN: I built AI Agents for continuous customer feedback

Hacker News - Fri, 04/11/2025 - 4:11am

Growth is harder than ever. Acquisition costs, churn, customer expectations, # of competitors—all way up! Understanding your customer is your moat but traditional methods aren't great.

Survey's lack depth. Human interviews aren't repeatable, scalable or convenient (for you or your customers).

Franko let's you create and deploy agents to have short, targeted conversations at critical points in your customer journey.

Configure your agent with a "Conversation Plan" and other details (i.e. # questions to ask, etc) and then send directly to customers!

I.e. you might configure an agent to understand customer churn, and send your relevant agent link in an automated churn email sequence.

Customer clicks, opens a ChatGPT-like UI, and the agent guides them from there! Short, sharp, insightful!!

Let me know what you think :)

Cheers, Fletcher

Comments URL: https://news.ycombinator.com/item?id=43651518

Points: 1

# Comments: 0

Categories: Hacker News

Show HN: I made a Chrome extension that turns each new tab into a birding moment

Hacker News - Fri, 04/11/2025 - 4:07am

I built BirdTab, a Chrome extension that introduces you to a different bird species each time you open a new tab. It features stunning photos, bird calls (with a mute option), and interesting facts to bring a moment of nature into your browsing experience.

A year ago, I moved to the Himalayan foothills and was captivated by the song of a Blue Whistling Thrush outside my window. This sparked a fascination with birds. As a developer, I still spent hours in front of Chrome tabs, so I created BirdTab to bring a slice of that birding experience into my daily routine.

Comments URL: https://news.ycombinator.com/item?id=43651498

Points: 1

# Comments: 0

Categories: Hacker News

Show HN: Open-source, cross platform document data extraction with no OCR

Hacker News - Fri, 04/11/2025 - 4:02am

The traditional pipeline for unstructured data extraction typically follows these steps: 1. Image → OCR Model (e.g., Google Vision) → Layout Model (e.g. Surya) → LLM → Final Answer However, this can be streamlined using a Vision-Language Model (VLM): 2. Image → VLM → Final Answer

Recently VLMs have improved a lot for OCR and document understanding tasks, specifically the Qwen-2.5-VL series. We can run the Qwen-2.5-VL-7B-AWQ model locally with just 16GB VRAM, and perform end-to-end information extraction (fields and table extraction) without any external models.

Hallucination with VLMs One question I am asked often is, higher hallucination with VLMs compared to OCR model. This is a valid point. But, even with correct OCR and Layout formatting, LLM can still hallucinate, and can give incorrect final answers. Layout models often struggle with complex documents (e.g., tables, complex sparse document). If the formatted text from the layout model is incorrect the LLM model will always produce incorrect extraction with high confidence.

Check out our GitHub repo for implementation details:

GitHub: https://github.com/NanoNets/docext

Comments URL: https://news.ycombinator.com/item?id=43651477

Points: 2

# Comments: 0

Categories: Hacker News

Pages