Tallha Ijaz

Tallha Ijaz

Machine Learning, Systems and Control

About

I am a Machine Learning Engineer with a background in Electrical Engineering and Project Management, along with academic and professional experience internationally. I have built a strong theoretical foundation in classical ML and advanced DL, which I have complemented with practical internships in Sweden and a master's thesis and an exchange semester in Germany. I am skilled in Python, PyTorch and TensorFlow for ML model development, deployment and optimization for tasks in and around Computer Vision, Robotic Perception, 3D Reconstruction and Applied AI.

Badges

Tastemaker
Tastemaker
Gone streaking
Gone streaking
Gone streaking 5
Gone streaking 5

Forums

We removed our pricing page for one week and our sales increased.

The structure of our pricing page included three tiers, a comparison table consisting of fourteen rows, and a FAQ section underneath it where questions that nobody ever asked were answered. The conversion rate was 1.9% for five months.

Last month, I changed everything completely by reducing the number of tiers to one, changing the table into one paragraph, and putting there a sign-up button. This was supposed to be a temporary change, until I reconstructed the table. In this case, the conversion rate reached 3.4% and maintained at this level.

15d ago

Building real-time state and checkout webhooks using Gemini and Antigravity with zero CS background

Hey everyone,

I don't come from a formal engineering background, but I've been experimenting with how far pure natural language prompting can push a functional full-stack app.

I recently shipped an interactive daily bidding leaderboard where users displace ranks dynamically, resetting every midnight UTC.

18d ago

Vibe coding is fun until the AI becomes dumb all of a sudden.

I used a few different tools to vibe code my saas. I used lovable, chatgpt, gemini to code my saas. The thing is, there is a huge problem of these AI's of forgetting when you chat for a longer time.

This has happened to me a lot of times. One solution I found is that you can start a new chat to overcome this problem. But, here the problem arises to explain the AI all the context which you had in the previous conversation.

Curious, has anyone else ever ran into this problem? How did you manage to overcome this?

View more