At Spearhead, we believe AI is not just a tool—it’s an essential part of how we think, build, and connect. During our recent Office Hours with the ABC (AI BootCamp) cohort, we shared a glimpse of the AI tools we use day-to-day. The list sparked interest and conversation, and we realized we had only scratched the surface.

‍

Here’s a detailed look at our full AI and tech stack, from code to concept to conversation:

‍

1. OpenAI ChatGPT o3 — Our primary deep research assistant. Whether summarizing papers or testing product hypotheses, ChatGPT cuts down our analysis time drastically.

2. ChatGPT / Google Gemini / Grok (Image Gen) — We alternate these depending on the project for generating visuals and storyboards to prototype quickly.

3. Anthropic Claude 4 — Exceptional for debugging and long-context reasoning. We use it for technical deep dives and logic-heavy workflows.‍

4. Granola Pro — The only AI note-taking tool we recommend. It captures context from meetings and delivers clear, actionable summaries.

5. Perplexity Enterprise Pro — Our go-to for rapid research. The new Perplexity Labs functions help us build custom agents tailored to client needs.‍

6. Custom RAG & GraphRAG using NVIDIA NeMo — This is where IP becomes product. We use LangChain, FAISS, Dgraph, Neo4j and more to power domain-specific retrieval applications.

7. Lovable — Turns design ideas into visual prototypes in under 20 minutes. Speeds up feedback loops significantly.

8. Replit — Instant coding environments let us build and test microservices quickly. Great for early-stage POCs.

9. Cursor — Inline AI coding and project navigation. Makes pull requests faster and improves codebase comprehension.

10. Firebase Studio — Ideal for building scalable backends. We use it especially for AI dashboards and internal tooling.

11. n8n — Our event orchestration layer. Automates workflows, integrates systems, and eliminates repetitive manual tasks.

12. Vercel — Streamlines front-end deployment and UI testing. Our frontend team loves it for how fast it moves.

13. Wispr Flow — Voice-first input system. We use it to craft long AI prompts and speed up email/text responses.

14. Meeting People IRL (in real life) — Not a tool, but possibly the most important part of our stack. Nothing beats face-to-face for clarity, alignment, and creative breakthroughs.

15. Gamma and Napkin AI — fast slide deck creation for internal use.

16. Hyperscalers — AWS Bedrock, Azure, Google Cloud, CoreWeave depending on client preferences.

17. Open-source LLMs — Meta Llama, Mistral AI for specialized or cost-optimized builds.

18. Bizagi — process automation and workflow logic.

19.Databricks, Snowflake, Datadog — because great AI starts with great data.

20. NotebookLM — brilliant for summarizing lengthy content and generating podcast outlines.

21. Other notables — Microsoft Office, Airtable, Notion, Webflow: still indispensable.

‍

Why it matters: This stack isn’t just about being tech-forward. It’s about accelerating impact. From first draft to final product, from a prompt to a pilot feature—these tools help us move faster and with greater precision.

‍

So that’s our list. Now we’re curious—what’s yours?

‍

FAQs

‍

1. Why do you use multiple LLMs like ChatGPT, Claude, and Perplexity?
Each LLM has strengths: ChatGPT excels at structured outputs and plugins; Claude handles longer context windows and nuanced reasoning; Perplexity is unbeatable for sourcing and citation-backed research. Using them together helps us triangulate insights faster.

2. What makes Granola Pro your go-to note-taking AI?
Granola Pro captures meetings with high transcription accuracy, labels decisions automatically, and integrates with project management tools — it’s more reliable than alternatives we’ve tested.

3. Why is “meeting people IRL” part of an AI stack?
Because no AI tool can replicate human chemistry, real-time feedback, or the creative energy of face-to-face conversations. It’s a reminder that human-first workflows still outperform, especially in strategy or early-stage product ideation.

‍