Hire Me
If you’re running production AI and something’s quietly broken, I can help. Two engagement shapes:
Production AI Audit
A focused review of your production AI deployment against the 35-pattern failure catalog (Field Reference — bundled with the suite). Findings + corrective-action plan delivered within the engagement window.
| Tier | Scope | Investment | Timeline |
|---|---|---|---|
| Audit Lite | One system, top-5 findings, written report | $1,500 | 1 week |
| Audit Standard | Full audit, all 14 patterns, 5 Cs findings, 90-day follow-up | $3,000 | 2-3 weeks |
| Audit + Workshop | Standard + 2-day team workshop + first monthly audit included | $7,500 | 3-4 weeks |
What you get:
- Written report against the 35-pattern catalog (which patterns are present, which aren’t, severity)
- Specific corrective actions (not “implement monitoring” — exact code changes, hooks, checks)
- 90-day follow-up call (Standard + Workshop tiers)
Email: hello@temhan.dev with subject AI audit inquiry. Include 1 paragraph on your stack + which tier you want.
Custom MCP Build
Adapt the 7 MCP servers I’ve shipped — or build a new server entirely — for your specific stack. Different agent harness, custom logging schema, AWS Bedrock metering, vendor-specific billing API, internal tools.
| Tier | Scope | Investment | Timeline |
|---|---|---|---|
| Simple | Custom claim-extraction prompts + tuned thresholds for your domain | $8,000–$10,000 | 1-2 weeks |
| Standard | Multi-language scanners (TypeScript / Go / Rust AST walks), custom severity rules, integration with your CI | $15,000–$25,000 | 2-4 weeks |
| Complex | LLM-as-judge backend with hosted model, persistence, full CI integration, audit-trail | $30,000–$45,000 | 4-8 weeks |
What you get:
- Working MCP server tested against your real production data (not synthetic fixtures)
- Source code transferred to your repo (not hosted on my infra)
- Documentation + setup walkthrough + 30-day post-ship support
- Optional: open-source the generic version, keep the domain-specific layer private
Email: hello@temhan.dev with subject Custom MCP Build inquiry. Include 1 paragraph on your stack + which tier.
Advisory / short-form consulting
For one-off questions, ad-hoc reviews, or pre-engagement scoping — happy to do a 60-min advisory call. Hourly rate, prorated.
Email: hello@temhan.dev with subject Advisory.
How engagements work
- You email with a one-paragraph description of your stack + the tier you’re considering
- I reply within 2 business days with a 30-min discovery call slot
- Discovery call confirms scope, timeline, and any red flags. Free.
- One-page SOW sent within 24 hours of the call. You review, send back any redlines.
- Signed SOW + first payment triggers start. I never start work pre-payment.
- Engagement runs with weekly check-ins.
- Deliverables hand-off + 30 or 90-day follow-up depending on tier.
Average from first email to signed SOW: ~5 business days.
What I don’t do
To be honest about scope mismatch:
- No fixed-price engagements over $45k. Above that, hourly is safer for both of us.
- No subscription / retainer engagements until we’ve completed at least one fixed-scope engagement together.
- No “rescue” work without a paid audit first. If you say “we deployed AI, it failed, please fix it” — the right first step is the audit, not skipping straight to fix-it.
- No work without signed SOW + first payment. Verbal “let’s get going” = unpaid work. Even with good-faith clients.
Common questions
“What stacks have you worked with?”
Production AI deployments across LangChain, LlamaIndex, vanilla SDK calls, custom agent harnesses, OpenClaw, Claude Code, Cursor, Anthropic API, OpenAI API, Gemini, AWS Bedrock. The patterns generalize across all of them — that’s the point of the 35-pattern catalog.
“Do you have references?”
Engagement-specific. Mention in your inquiry email if references matter for your buying process; happy to share once we’re past discovery.
“What if my problem isn’t on the catalog?”
Then we add it. The catalog is a working artifact, not a fixed bible.
“Can you do a free audit so I can decide?”
The free version is the AI Production Auditor Custom GPT — paste your config, get a 5 Cs report. Same methodology, no signup beyond ChatGPT. Useful as a self-serve first pass before commissioning paid work.