The SME Standard exists because the AI-data industry trains nobody — and we were the ones grading the consequences. What began as internal onboarding guides for fellow experts on frontier-lab projects became the professional school this industry never had.
Frontier AI is trained and evaluated on the judgment of human experts. That judgment deserves a craft, a curriculum, and a credential. We build all three.
Credentialed professionals deserve better than unpaid assessments and rejection without feedback. We teach the craft the platforms screen for — openly, rigorously, and with honest market data.
Vendors and labs deserve expert benches that pass audit the first time. We convert applicant-pool waste into yield, measured in approval-rate lift — not completion certificates.
A vendor-neutral standard — one published integrity code, one credential, one bar for what "expert-grade" means in AI training data.
Wasim is a corporate attorney and management consultant with over two decades of cross-border practice spanning M&A, venture capital, and financial services — including more than $7 billion in advised transactions and oversight roles across 100+ companies.
For the past several years his practice has centered on the AI frontier: designing evaluation frameworks for frontier legal LLMs and serving inside major AI training-data programs in every seat that matters — task creator, RLHF quality-control reviewer, benchmark auditor, and SME trainer. He has authored expert-level evaluation tasks rated at the highest complexity tiers, built and audited grading rubrics with 100+ weighted criteria, and written the onboarding guides used to bring other legal professionals onto live frontier-lab projects.
That last role is where The SME Standard began. The training materials he built for colleagues — rubric explainers, QC checklists, plain-language guides for brilliant lawyers who had never seen an annotation interface — worked. The school makes that standard available to everyone.
Wasim is also the founder of Fidem Praxis, an advisory firm at the intersection of counsel, strategy, and AI for regulated industries.
The SME Standard is built as a practitioner institution, not a personality. Instruction, grading, and audit calibration are delivered by a bench of working subject-matter experts — attorneys and finance professionals active on live AI-data programs — supported by guest reviewers and advisors across law, finance, and AI evaluation. As cohorts graduate, our best alumni join the bench as teaching assistants and certified reviewers: the school staffs itself the way the industry does, on demonstrated craft.
Every lab is led by someone doing the work now — current project experience, current standards, current failure modes.
Domain specialists who grade capstones blind and keep our calibration honest across legal and finance tracks.
Top graduates return as TAs and certified reviewers — inter-rater tested, paid per package, the reviewer model the industry itself uses.
Instructor project work sits under confidentiality obligations, which we honor completely. Our curriculum is clean-room: public sources and original synthetic examples only.
We teach the skill, never the test. No platform assessment content, no impersonation coaching, no NDA breaches — ours or anyone's.
Sourced pay data, published downsides, no income guarantees. Our buyers are professionals; we write like it.
Cohorts are graded against published rubrics; outcomes are tracked and reported. If we can't measure a claim, we don't make it.