Location: India (H-1B)
  Remote: Yes
  Willing to relocate: Yes
  Resume: https://abhinavkulkarni.github.io/assets/Resume.pdf
  Technologies: Machine Learning, NLP, LLM, Generative AI
  Email: abhinavkulkarni[at]gmail[dot]com

I have a keen eye for product development. 12+ years of experience working for SV mid/large-sized companies.

While my focus is Machine Learning, I'm a swiss army knife in overall backend development, System Design, and dealing with scale.

My LinkedIn is the bio.

Location: India (H1-B approved)

Remote: Yes

Willing to relocate: Yes

Machine Learning engineer with expertise in Deep Learning focussed on NLP and LLMs. I have experience dealing with web-scale data.

I recently helped integrate AWQ 4-bit quantization to HuggingFace's TGI: https://bit.ly/3QuZmKk

I have a keen eye for product development. 12+ years of experience working for SV mid/large-sized companies.

While my focus is Machine Learning, I'm a swiss army knife in overall backend development, System Design, and dealing with scale.

My LinkedIn is the bio.

  Location: India (H-1B)
  Remote: Yes
  Willing to relocate: Yes
  Resume: https://abhinavkulkarni.github.io/assets/Resume.pdf
  Technologies: Machine Learning, NLP, LLM, Generative AI
  Email: abhinavkulkarni[at]gmail[dot]com

I recently helped integrate AWQ 4-bit quantization to HuggingFace's TGI: https://bit.ly/3QuZmKk

I have a keen eye for product development. 12+ years of experience working for SV mid/large-sized companies.

While my focus is Machine Learning, I'm a swiss army knife in overall backend development, System Design, and dealing with scale.

My LinkedIn is the bio.

Machine Learning engineer with expertise in Deep Learning focussed on NLP and LLMs. I have experience dealing with web-scale data.

I recently helped integrate AWQ 4-bit quantization to HuggingFace's TGI: https://bit.ly/3QuZmKk

I have a keen eye for product development. 12+ years of experience working for SV mid/large-sized companies.

While my focus is Machine Learning, I'm a swiss army knife in overall backend development, System Design, and dealing with scale.

My LinkedIn is the bio.

Just like many similar methods, this is based on logit biasing, so it may have an impact on quality.

HyDE sounds like an interesting approach. All dense retrieval approaches suffer from the problem you outlined in the blogpost. Have you looked at keyword-based or late-interaction models for retrieval such as ColBERTv2[1]? I find that late-interaction methods seem to offer best trade-off between semantic intelligence (precision) and retriving relevant documents (recall).

[1] https://github.com/stanford-futuredata/ColBERT