Equity Only | Pre-Seed Stage Startup | India (Preferred, Remote-Friendly)
About Us
We're building a Robinhood-class mobile app powered by AI that turns massive, messy data into actionable trade ideas and auto-executes via users connected brokers. We are a startup based in the US and have a registered office in India. We need your investment of a minimum of 20 Hours per Week (Part-time), which offers significant returns through equity, as you continue working while we work to secure funding and onboard you as a Full-time employee within the next 6-8 months.
You'll be a partner with our AI engineers, quantitative research team, and frontend and backend teams.
The Role
You will lead LLM model selection and adaptation for financial applications, comparing open- and off-the-shelf models (e.g., the OSS, Llama 4 family, Mistral, DeepSeek) on our tasks and data.
You'll design a decision framework for RAG vs. fine-tuning vs. full/continued pretraining, and implement efficient methods (e.g., LoRA/QLoRA) to hit latency, quality, and cost targets. You'll also stand-up objective evaluations (HELM/MTEB-style) and production guardrails appropriate for a fintech app.
What You'll Do
- Model selection & benchmarking
- Compare open/off-the-shelf LLMs on finance tasks, build scorecards (quality, latency, context, cost), and run holistic evaluations inspired by HELM and MTEB.
- Adaptation strategy: Fine-tuning vs. pretraining
- Propose a decision tree for when to use fine-tuning (stable domain tasks) vs. continued pretraining (deep domain needs).
- Justify trade-offs for cost, safety, and maintenance.
- Efficient fine-tuning & training
- Implement LoRA/QLoRA pipelines.
- Quantify gains vs. full fine-tune.
- Document computing/memory budgets and inference impacts.
- Finance-aware modeling
- Evaluate domain LLMs and datasets (e.g., FinGPT, BloombergGPT) and codify finance-specific evaluation sets (tickers, filings, news, risk language).
- Own a benchmark pack (e.g., FinanceBench, FinQA, FiQA, and, if relevant, FinBen tasks) with pass/fail gates and a living leaderboard
- Risk, reliability & governance
- Define offline/online evals, hallucination tests, and failure modes for advice/explanations; partner with compliance on clear disclosures consistent with Robo-adviser guidance.
- Prod integration
- Ship models with tracing, prompts/versioning, and A/B or champion-vs-challenger evaluation.
- Monitor latency/cost/quality SLOs in production.
What We're Looking For
- 8 –12+ years in data science/ML with strong Python (PyTorch/Transformers) and LLM ops.
- Hands-on with LLM evaluation, prompt/program design, and RAG stacks; proven LoRA/QLoRA experience.
- Prior bake-offs among open-weight models (e.g., Llama/Mistral/DeepSeek) with scorecards (quality, latency, context, $/1k tokens). (Seen across fintech/LLM roles.)
- Experience choosing between fine-tuning and continued pretraining, with clear cost/quality trade-offs.
- Finance domain familiarity (equities/options) strongly preferred; ability to craft domain evals (EDGAR/filings, news, corporate actions).
