§ Advisory
Work with me
I take on a small number of engagements alongside my full-time work at JuiceLabs. Typical scope is 2–6 weeks and covers three tracks.
Track 01
AI adoption strategy
For teams trying to figure out where Claude or other LLMs actually make sense in their workflow, and where they don't. Architecture review, prompt engineering patterns, evaluation strategy. I run weekly Claude Code sessions for our 20-person team and consult directly with enterprise clients at JuiceLabs on exactly this, so the frameworks are already production-tested.
Track 02
Agentic systems design
If you're building an agent workflow and want a review before committing to an architecture. LangGraph versus plain chains, evaluation loops, failure-mode planning, cost and latency tradeoffs. I've shipped production agentic systems generating creative at scale. The scars are real.
Track 03
AI builds and prototypes
Focused builds of AI-powered web applications, agentic workflows, and MVPs. Full-stack, end to end: frontend, backend, model orchestration, evaluation. Best fit is 4–6 week scopes with a clear demo at the end. Good examples of what I ship are linked on the home page: Agentic RAG, AgentVC, Policy Compare.
Not a fit
- Multi-month full-time contracts (I'm full-time at JuiceLabs)
- Body-shop staff augmentation or team placements
- Crypto or speculative use cases
§ Send an inquiry
Tell me what you’re working on
What you’re building, your timeline, and the specific decision you’re trying to make. If it’s a fit I’ll reply within a few business days.
Prefer email?
dhruvpuri00766@gmail.com