Job description

Problem Statement-Solution

Only 10% of India speaks English and 90% speak over 25 languages and 1000s of dialects. The internet has largely been in English. A good part of India is now getting internet connectivity thanks to cheap smartphones and Jio. The non-English speaking internet users will balloon to about 600 million users out of the total 750 million internet users in India by 2020. This will make the vernacular segment one of the largest segments in the world - almost 2x the size of the US population. The vernacular segment has very few products that they can use on the internet.

One large human need is that of sharing thoughts and connecting with people of the same community on the basis of language and common interests. Twitter serves this need globally but the experience is mostly in English. Theres a large unaddressed need for these vernacular users to express themselves in their mother tongue and connect with others from their community. Koo is a solution to this problem.

About Koo

Koo was founded in March 2020, as a micro-blogging platform in both Indian languages and English, which gives a voice to the millions of Indians who communicate in Indian languages.

Currently available in Assamese, Bengali, English, Hindi, Kannada, Marathi, Tamil and Telugu, Koo enables people from across India to express themselves online in their mother tongues. In a country where under 10% of the population speaks English as a native language, Koo meets the need for a social media platform that can deliver an immersive language experience to an Indian user, thereby enabling them to connect and interact with each other. The recently introduced ‘Talk to Type’ enables users to leverage the voice assistant to share their thoughts without having to type. In August 2021, Koo crossed 10 million downloads, in just 16 months of launch.

Since June 2021, Koo is available in Nigeria.

Founding Team

Koo is founded by veteran internet entrepreneurs – Aprameya Radhakrishna (CEO, Taxiforsure) and Mayank Bidawatka (Co-founder, Goodbox & Coreteam, redBus).

Technology Team & Culture

The technology team comprises sharp coders, technology geeks and guys who have been entrepreneurs or are entrepreneurial and extremely passionate towards technology. Talent is coming from the likes of Google, Walmart, Redbus, Dailyhunt. Anyone being part of a technology team will have a lot to learn from their peers and mentors. Download our android app and take a look at what we’ve built. Technology stack compromises of a wide variety of cutting-edge technologies like Kotlin, Java 15, Reactive Programming, MongoDB, Cassandra, Kubernetes, AWS, NodeJS, Python, ReactJS, Redis, Aerospike, ML, Deep learning etc. We believe in giving a lot of independence and autonomy to ownership-driven individuals.

Problems we Solve

ML@Koo is a small and focused team with the following main objectives:

1. Make Koo a clean and safe platform.
2. Recommend content and "people to follow" to users they might be interested in.
3. Make content more discoverable by extracting structures in it, such as topics.
4. Each of the above objectives are open problems, when combined with challenges like people being creative while being mischievous and working with Indic languages where not much NLP work has been done. We use the latest technologies to solve these problems and meanwhile build our own state-of-the-art algorithms.

Job Description

• Design, develop, test, deploy, maintain and improve ML models/infrastructure and software that uses these models.
• Manage individual project priorities, deadlines and deliverables.
• Contribute to engineering efforts from planning and organization to execution and delivery.
• Design and build new data pipelines from scratch, all the way to deploying them to production
• Build models and algorithms to understand user interest, user intent, and improve content relevancy.
• Improve existing recommendation systems, experiment with new directions, and provide ML solutions in recommendation systems.
• Mentor junior ML engineers in their growth and project deliverables.

Requirements:

• Minimum 4 years’ experience in the field of Machine Learning.
• Proficiency in Python, SQL
• Strong theoretical foundations in ML.
• Experience working with ML and DL frameworks (Scikit-learn, Tensorflow, PyTorch, Keras)
• Experience with AWS infrastructure (Sagemaker, S3, EKS etc.)
• Experience in NLP/NLU
• Comfortable with frameworks like MLFlow, AirFlow, Git, Flask, Docker, Spark

Technology skill sets required for a matching profile

1. Building and deploying ML models.
2. AWS infrastructure management.
3. Constructing, evaluating, and tuning ML Pipelines.
4. Spark, python, MLFlow, AirFlow, Git, SQL