01
Security & reliability of an LLM or agentic system
I take your system as it actually runs and look for where it breaks — not in theory, in your code.
- OWASP ASI Top 10 mapped to controls actually in place, gaps named
- Adversarial review of the data path: inputs, memory, tools, outputs
- Supply chain and machine-identity audit
- Prioritized P0 / P1 / P2 report, each with its fix and its cost
- CI gates ready to wire in, so a closed hole stays closed
Fixed fee — on request3–5 days
Read →
02
On-premise LLM stack, 100% local
The reference architecture behind this portfolio, fitted to your hardware and your regulatory constraints.
- Resource-aware inference gateway: several models, one RAM budget
- RAG where every answer is sourced — and which refuses rather than invents
- Tool-using agents with explicit approval on anything that writes
- Supervised services, health endpoints, privacy-respecting usage journal
- Knowledge transfer: your team takes over, documentation included
Day rate — on requestengagement
Read →
03
Continuous eval, CI gates, observability
An AI system degrades silently. Mine is watched every night; yours can be too.
- Nightly eval harness, pinned baselines, explicit drift budget
- Vulnerability and version watch, exceptions tracked by name
- Observability: cold and warm latencies, refusal rate, real cost
- Monthly report readable by a steering committee, not only by an engineer
- Light on-call on regressions caught by the gates
Monthly — on requestrecurring
Read →