ML/LLM Engineer

We are a venture capital fund based in Singapore, investing in high-growth startups across Southeast Asia. As part of our data-driven investment strategy, we are building proprietary tools to gain deeper insights into people, companies, and talent networks. We’re now looking for a Machine Learning / LLM Engineer to help us harness the power of large language models (LLMs) and unstructured people-centric data to drive intelligent deal sourcing and talent discovery.

Role Overview: You will work directly with the investment and technology teams to develop and deploy LLM-based tools that can process, analyze, and extract insights from large-scale people-centric datasets — including LinkedIn profiles, resumes, public databases, and unstructured sources. This role is ideal for an engineer with a hacker mindset, strong ML/LLM experience, and a deep curiosity about talent signals and people data.

Responsibilities:

• Design and implement ML/NLP pipelines using LLMs to extract structured insights from LinkedIn-like data and resumes.

• Build tools and internal APIs to power investor workflows for talent intelligence and deal sourcing.

• Fine-tune or prompt-engineer open-source and proprietary LLMs for use cases such as profile summarization, skill inference, and network mapping.

• Work with unstructured data sources and apply entity recognition, linking, and classification models to enrich data quality.

• Collaborate with investment team members to productize insights and iterate quickly based on feedback.

• Stay updated on the latest developments in LLMs, vector search, and embeddings to ensure state-of-the-art performance.
  

Requirements:

• 3–6 years of experience in machine learning, with strong focus on NLP and LLMs (e.g., GPT, LLaMA, Mistral, Claude).

• Experience working with people- or LinkedIn-centric datasets (e.g., resumes, professional bios, talent graphs).

• Strong Python skills and experience with ML libraries (e.g., Hugging Face Transformers, LangChain, OpenAI APIs).

• Comfortable working with unstructured data and applying techniques like embeddings, entity extraction, and prompt engineering.

• Familiarity with vector databases (e.g., FAISS, Pinecone, Weaviate) and scalable data pipelines.

• A hacker mindset — fast, resourceful, and comfortable in early-stage environments.

• Bonus: Experience with graph-based models or social/professional network analysis.
  

What We Offer:

• Opportunity to shape the data and tech stack within a forward-thinking VC fund.

• Close collaboration with investors and partners on high-impact internal tools.

• Competitive compensation and access to cutting-edge LLM research and infrastructure.

• A fast-moving, entrepreneurial environment with regional impact.