AI/ML engineer (with Backend skills)

This is a key position with us for someone who has strong engineering fundamentals, wish to build along with us in a fast-pace environment and has worked in AI.

This role is for a AI/ML engineer who has worked on Backend previously.

Application form: https://forms.gle/5c99AR1RFVNxz9WS9

Requirements:

• Background in AI/ML
• 1.5+ yrs of experience in AI/ML and built related tech (Python/NodeJS)
• Working knowledge on - LLMs, RAG, vector DBs, etc.
• Understand quality implementations
  • Eg - improving retrieval in RAG, differences in training techniques likes QLoRA vs QA-LoRA..
• 1+ yr of experience in Backend engineering
  • Experience with a relational database - PostgreSQL/MySQL
  • Worked on NodeJS/NestJS/Django building an application/product
  • Understanding of rate limiting, scale, basics of security

Responsibilities:

• Leading some core initiatives at SuperKalam
• Building custom LLM models - fine tuning, RLHF, RLAIF
• Running AI experiments and set up pipelines
• Research and implement appropriate approaches and tools.
• Keep self updated with evolving technology - say, actively following hugging face, training techniques models, tooling?
• We thrive on smooth communication, ownership and agility. It’s a must.

Puzzle fit:

We are an early-stage startup, where we have a good product with solid roadmap, lot of users, an impactful vision and a good runway.

If you are someone who is = UNION(High learning agility, Strong ownership, a long term player, Outcome oriented, High degree of candour), you are a 🧩 for us. And ofcourse, wants to make an impact worth remembering in the world.

You resume will be short listed as per your skills and prior experience. After this, an introductory call will be set.

Introductory call (20 min) - We will discuss your work in detail. You can ask relevant questions. Depending upon the depth of skillset, the next steps will be discussed.

Round 1 (Optional): Assignment - 2 days. This is to evaluate your scientific approach to the problem statement and solutioning. We will have 30min discussion on feedback - Skipped if your previous experience is substantial.

Round 2, F2F, 1.5 hrs. This round is for fundamental programming and LLD system design.
Round 3, F2F, 1hr. This round will have ML design and architecture problem statements.

Final discussion round