/consulting
Selective consulting. Fixed-scope. Real code.
Sadaf Labs takes 2 client projects per quarter— the ones where we'd use our own lab tools (PolicyLint, Trace Replay, Eval Harness) to ship faster. Every engagement starts with a free 20-min scoping call.
Agent Trust Audit
1–2 weeks · from $5k
We map your live agent, find policy gaps and prompt-injection surfaces, and ship a hardened system prompt + audit-log scaffold.
- PolicyLint pass on your system prompts
- Tool-scope review (least privilege)
- Audit log + replay scaffold
- Red-team prompt set you can reuse
RAG Knowledge System
3–4 weeks · from $12k
Search and Q&A over your unstructured docs with citations and access control.
- Document ingestion pipeline
- Vector store + hybrid retrieval
- Citation-grounded answers
- Per-team permissions
Multi-Agent Build
4–8 weeks · from $25k
Supervisor + specialist agents for a real workflow, with policy gates, human-approval steps, and observability from day one.
- Custom agent design (3–6 agents)
- Tool integrations (CRM, DBs, APIs)
- Policy + audit + replay
- Human-in-the-loop approvals
AI MVP in 30 days
4 weeks · from $20k
Idea → working product → first paying user. One sprint, fixed scope, real code.
- Product scoping workshop
- Full-stack MVP (Next.js + AWS)
- AI core (LLM + retrieval)
- Launch + analytics
Fractional AI Lead
Monthly retainer · from $8k/mo
Embedded AI engineering leadership for your team — architecture, hiring, and code where it matters.
- Weekly architecture reviews
- Code contributions where it matters
- Vendor / tooling decisions
- Hiring & interviewing AI talent