Unsiloed AI — Founding ML Researcher

Unsiloed AI — Founding ML Researcher

Type: Full-time | On-site | San Francisco, CA Compensation: $200,000–$300,000 + 0.1%–1% equity Hiring count: 1 Visa sponsorship: Yes — H-1B, O-1, OPT Reports to: Founders (co-founder Aman)

About Unsiloed AI

Unsiloed converts complex documents into LLM-ready data. Its production-grade APIs ingest, parse, structure, and split documents across formats (PDF, PPT, Excel) into clean Markdown and JSON that AI agents and LLMs can work with reliably from day one. The platform preserves document structure and hierarchy while capturing domain-specific context critical for high-stakes workflows across finance, legal, and healthcare.

Founded: 2024 | Team size: 1–10 (Seed) | Total funding: not confirmed on role page Industry: AI Tools / Document AI Website: https://www.unsiloed.ai/ Office: San Francisco, CA

Why Candidates Should Join

• Define the technical ceiling: This role sets the engineering/research archetype for a small, talent-dense founding team where the first few hires matter disproportionately.
• End-to-end ownership: Own the full ML lifecycle — research experimentation training evaluation production deployment — with full autonomy to drive systems from idea to production.
• Direct founder access: Work directly with the founders to shape product direction and long-term ML/engineering strategy.
• Real production impact: Ship document-AI models used in high-stakes finance, legal, and healthcare workflows.
• Perks: Competitive comp + equity, full health insurance coverage, free DoorDash/Uber credits, free company laptop.

Intake Call Summary

• Not available — the role page includes an Intake Video but no transcript or written summary.

The Role

Founding ML Researcher shaping Unsiloed's ML research direction and translating cutting-edge research into production-ready document-AI models. Research-heavy but product-oriented — comfortable moving between theory, experimentation, and real-world deployment.

What You'll Be Doing

• Own the end-to-end ML lifecycle: research experimentation training evaluation production deployment
• Work with the engineering team to transition research into deployed, scalable systems
• Drive best practices for data, experimentation, evaluation, and model iteration
• Work directly with the founders to shape product direction and engineering strategy

Tech stack: Not specified on the role page. Work centers on VLM-based document understanding, layout models, document parsing of unstructured data, and production model serving / inference optimization.

Requirements

• Training and deploying state-of-the-art models for parsing and understanding unstructured data
• Experimenting with novel techniques to improve layout models and VLM-based document understanding
• Building data pipelines, evaluating model performance, and integrating models into production systems

Green Flags

• Experience at top research labs like DeepMind, OpenAI
• Strong publication record in computer vision/ML
• Hands-on implementation and production experience

Red Flags

• Candidates whose recent focus has been management, rather than individual contributor work
• No hands-on experience training AI models

Role Details

Salary$200,000–$300,000Equity0.1%–1%ExperienceAny (new PhD grads ok)On-site policyOn-site — San Francisco, CAVisa sponsorshipH-1B, O-1, OPTEmployment typeFull-timeLocationSan Francisco, CA

Screening Questions

• None provided on the role page.

Interview Process

Stage 1 — Pending Approval — Candidates awaiting initial approval. Stage 2 — Initial chat with co-founder Aman Stage 3 — Technical Coding Round (1 hour) — Claude Code/Codex-assisted round solving a small, hard problem. Not DSA/Leetcode. Stage 4 — 1-week work trial (flexible scheduling) — Compensated; flexible on timing/weekends to accommodate current commitments. Stage 5 — Offer Extended Stage 6 — Candidate Hired — Candidate accepts and starts.

Ideal Companies & Backgrounds

Updated: date not shown on role page Top research labs / frontier AI — Google DeepMind, OpenAI, DeepSeek, Anthropic, XAI