Use whenever the user wants to find, shortlist, vet, or enrich US AI/ML/data consulting firms (consultancies) — AI/ML development, MLOps, generative AI / LLM apps (RAG, chatbots, agents), computer vision, NLP, recommendation systems, data engineering, BI/analytics. Triggers on "find an AI/ML consulting firm to build our recommendation engine", "shortlist three RAG/LLM consultancies for an enterprise chatbot", "compare three AI/ML consulting firms with strong ratings", or "pull contact info for these 8 AI consultancy domains", even when described indirectly (we want to use AI for X, deploy ML to production). Drives the ServiceGraph API (api.servicegraph.co) — a 100k+ US firm catalog filterable by industry, services, location, size, ratings. Defer to find-software-developer for general app/backend work where AI is just a feature. Skip in-house ML/data hires, LLM/AI-tool comparisons (ChatGPT vs Claude), "how do I fine-tune X" DIY questions, AI courses for individuals, non-US firms, individual freelancers.
Use whenever the user wants to find, shortlist, vet, or enrich US AI/ML/data consulting firms (consultancies) — AI/ML development, MLOps, generative AI / LLM apps (RAG, chatbots, agents), computer vision, NLP, recommendation systems, data engineering, BI/analytics. Triggers on "find an AI/ML consulting firm to build our recommendation engine", "shortlist three RAG/LLM consultancies for an enterprise chatbot", "compare three AI/ML consulting firms with strong ratings", or "pull contact info for these 8 AI consultancy domains", even when described indirectly (we want to use AI for X, deploy ML to production). Drives the ServiceGraph API (api.servicegraph.co) — a 100k+ US firm catalog filterable by industry, services, location, size, ratings. Defer to find-software-developer for general app/backend work where AI is just a feature. Skip in-house ML/data hires, LLM/AI-tool comparisons (ChatGPT vs Claude), "how do I fine-tune X" DIY questions, AI courses for individuals, non-US firms, individual freelancers.
Drive the ServiceGraph API (https://api.servicegraph.co) to find,
shortlist, and enrich US AI/ML and data consultancies via the
pro_services dataset. The catalog tags firms with
industry:data_ai_consulting and a 4-tag service sub-taxonomy:
ai-ml-development (the largest at ~12k firms), data-analytics,
cloud-services, and api-integration. There is no
data-engineering or business-intelligence sub-tag —
data-analytics covers both. Confirm exact tag names via
/v1/datasets/pro_services/fields?include_values=1.
Always pin industry:data_ai_consulting. This skill exists to
do that automatically — the user shouldn't have to think about catalog
taxonomy.
Any HTTP client works (curl, fetch, requests). Examples below use curl.
Sibling skills — defer when scope is different
General application or backend dev that uses AI as a feature
(e.g. "build us a SaaS with an AI chatbot tab") → find-software-developer.
Web/site projects that include some AI → find-web-developer.
AI-related marketing or content → find-marketing-agency.
This skill is for engagements where the AI/ML/data work IS the
deliverable.
When NOT to use this skill
Consumer AI courses or learning ("find me an online course to learn ML").
AI/LLM product comparisons ("ChatGPT vs Claude vs Gemini", "Cursor vs Copilot").
DIY/code tasks ("how do I fine-tune Llama", "review this PyTorch loop").
In-house ML/data hires (Machine Learning Engineer, Data Scientist).
Generic AI knowledge questions.
Non-US firms / individual freelance ML engineers.
Prerequisites
ServiceGraph access, either:
the ServiceGraph MCP server (https://mcp.servicegraph.co) loaded in
your harness — this plugin's .mcp.json wires it up; OAuth 2.1 + PKCE
keeps credentials in the harness sandbox — or
a ServiceGraph API key (vk_…, minted at
https://servicegraph.co/profile/api-keys) available as
SERVICEGRAPH_API_KEY in the environment or .env.local for the REST
path (setup steps under Auth below).
An HTTP client for the REST path — the examples use curl.
Instructions
The loop is free-first: discovery, validation, search, and brief reads cost
nothing; only unlock after the user confirms the spend.
Pick the call path — the ServiceGraph MCP tools if loaded, otherwise the
REST flow (MCP server and Auth sections below).
GET /v1/datasets/pro_services/fields?include_values=1 — confirm the
fields and values you plan to filter on exist.
Build the filter (Filter DSL below) and validate it with
GET /v1/datasets/pro_services/check — or draft it from plain English via
POST /v1/datasets/pro_services/translate-intent.
GET /v1/datasets/pro_services/search — present the free brief cards and
let the user pick.
Quote the unlock cost (10 credits per row, 30-day TTL) and get an explicit
go-ahead.
POST /v1/datasets/pro_services/unlocks with the chosen apexes; report the
revealed detail.
GET /v1/me/credits to report the remaining balance when asked.
MCP server (preferred for authed calls)
If your harness has the ServiceGraph MCP server loaded (tools
containing servicegraph), prefer those — OAuth 2.1 + PKCE keeps the
token in the harness sandbox. Otherwise use the REST flow below.
API surface (dataset id: pro_services)
Every endpoint requires the bearer (Authorization: Bearer vk_…).
No anonymous tier.
Endpoint
Cost
Use it for
GET /v1/datasets/pro_services/fields[?include_values=1]
free
Confirm data_ai_consulting industry value and sub-tag names.
GET /v1/datasets/pro_services/check?filter=…
free
Validate filter.
POST /v1/datasets/pro_services/translate-intent
free
{intent} → DSL filter + sanity count.
GET /v1/datasets/pro_services/search?filter=…&limit=
free
Brief firm cards + per-row unlock hint + total.
GET /v1/datasets/pro_services/:apex
free
One row brief; detail only if unlocked.
POST /v1/datasets/pro_services/unlocks
10 credits / firm
{apexes:[...]} ≤100; atomic; 30-day TTL on detail.
GET /v1/me/credits
free
Balance.
Cost model. Discovery / validation / search / brief reads are
free. Detail (url, phone, email, social, address, full platforms
map) costs 10 credits per firm and lasts 30 days.
Auth
vk_* API keys minted in the dashboard. Keep the token out of the
LLM context — never read .env* into your context; dispatch via
shell.
Try the call first through a shell wrapper that sources
.env.local:
On 401 prompt the user (don't accept the key in chat):
"Open https://servicegraph.co/profile/api-keys, create a
key, and add SERVICEGRAPH_API_KEY=vk_… to .env.local here
(or export it). Tell me when done. Please don't paste the key
into chat."
Retry after the user signals ready.
Filter DSL
GitHub-search-style.
filter := orExpr
orExpr := andExpr ("OR" andExpr)*
andExpr := notExpr (("AND")? notExpr)* # whitespace = implicit AND
notExpr := ("NOT" | "-") notExpr | atom
atom := "(" filter ")" | predicate
predicate:= IDENT op valueOrList | bareword
op := ":" | "=" | ">=" | "<=" | ">" | "<"
valueOrList := value ("," value)*
value := IDENT | NUMBER | tagAtEvidence
tagAtEvidence := IDENT "@" ("low"|"medium"|"high")
bareword := IDENT | NUMBER # → keyword:<bareword>
Four rules that bite: AND binds tighter than OR (use parens);
comma list = OR within one predicate; negation is -x or NOT x;
bareword = keyword search (quote multi-word phrases).
AI-flavored examples (validate yours with /check):
service_provided:data-analytics + keywords pipelines / engineering (no data-engineering tag)
BI / analytics
service_provided:data-analytics (covers BI too — no separate business-intelligence tag)
Cloud architecture for data/ML
service_provided:cloud-services
API / data integration
service_provided:api-integration
LLM apps / RAG / agents
llm, rag, agent (keywords)
Generative AI
"generative ai", genai
Computer vision
"computer vision", cv
NLP / IDP / document understanding
nlp, idp, "document understanding"
MLOps / model deployment
mlops, deployment
Recommendation systems
recommendation, recsys
Predictive analytics / churn / forecasting
predictive, forecasting, churn
Identifying firms — apex
Firms are identified by their apex domain (scaleai.com, not
www.scaleai.com/about).
Output
All responses are JSON.
Search returns free brief firm cards — apex, name, location, and
rating signals — plus a per-row unlock hint and the match total. Briefs
never include url, phone_primary, email_primary, legal_name,
address_full, or the full platforms map.
Unlock (POST …/unlocks) returns each unlocked firm's detail block —
contact fields, address, socials, the platforms map — plus per-item
billing; detail stays readable for 30 days.
Errors arrive as a JSON envelope
{"error": {"code": "…", "message": "…"}} — see Errors below.
Examples
A. AI/ML consultancy for a recommendation engine
User: "AI/ML consultancy to build our recommendation engine for an ecommerce site."
GET /v1/datasets/pro_services/search?filter=industry:data_ai_consulting+service_provided:ai-ml-development+recommendation+ecommerce&limit=10
# Present, get pick of 3. "Unlocking 3 = 30 credits, 30-day TTL."
POST /v1/datasets/pro_services/unlocks
{ "apexes": ["firm-a.com", "firm-b.com", "firm-c.com"] }
B. RAG / LLM consultancies for a chatbot
User: "Three RAG/LLM consultancies for an enterprise chatbot."
GET /v1/datasets/pro_services/search?filter=industry:data_ai_consulting+(rag OR llm)+chatbot+enterprise&limit=10
If thin, drop enterprise and surface client-tier signals from the
unlocked detail later.
C. Data engineering partner
User: "Data-engineering partner to build our analytics pipelines."
No data-engineering tag — data-analytics is the closest and
covers both BI and engineering. Pin the tag plus keyword:
GET /v1/datasets/pro_services/search?filter=industry:data_ai_consulting+service_provided:data-analytics+(pipelines OR engineering)&limit=10
D. MLOps for model deployment
GET /v1/datasets/pro_services/search?filter=industry:data_ai_consulting+mlops&limit=10
E. Indirect intent — "use AI to predict customer churn"
User: "We want to use AI to predict customer churn — who can help us build that?"
GET /v1/datasets/pro_services/search?filter=industry:data_ai_consulting+service_provided:ai-ml-development+(churn OR predictive)&limit=10
Or let the translator do the mapping:
POST /v1/datasets/pro_services/translate-intent
{ "intent": "AI consultancy to build customer churn prediction" }
F. Computer vision + healthcare vertical
GET /v1/datasets/pro_services/search?filter=industry:data_ai_consulting+"computer vision"+healthcare&limit=10
G. Quality threshold + Fortune 500 clients
GET /v1/datasets/pro_services/search?filter=industry:data_ai_consulting+service_provided:ai-ml-development@high+rating>=4+fortune&limit=10
"Fortune 500" as a structured filter isn't a thing — surface from
briefs or treat it as a keyword.
H. Custom LLM agent for customer service
GET /v1/datasets/pro_services/search?filter=industry:data_ai_consulting+(llm OR agent)+("customer service" OR support)&limit=10
I. BYO apex list — enrich domains
User pastes 8–20 AI consultancy domains:
GET /v1/datasets/pro_services/:apex per domain — free brief
(404 = not in catalog, no charge).
User picks N to fully enrich. POST /unlocks = 10×N credits,
atomic, detail returned.
Re-runs within 30-day TTL are free.
A 404 here often means the firm is actually a SaaS product company
(many AI vendors brand as "AI services" but operate as a product) —
filtered out of the catalog.
Gotchas
Always pin industry:data_ai_consulting. Without it, ai-ml-development as a service tag surfaces IT firms that list AI as a sub-service.
Defer to find-software-developer for general dev that uses AI as a feature. When the deliverable is a SaaS product or app and AI is one of several features, that's software-dev work; this skill is for engagements where AI/ML/data work IS the deliverable.
Catalog audit notes: AI/ML-tagged firms have a higher historical rate of misclassification (some are SaaS products, some are B2C ed-tech). If an unlock returns a SaaS product, flag and skip rather than recommend.
Many sub-niches are keyword-only. Multi-word sub-niches split into ANDed barewords unless quoted (computer vision → computer AND vision; "computer vision" → one phrase).
LLM-product comparisons (ChatGPT vs Claude vs Gemini) are NOT procurement — refuse.
AI courses for individuals (Coursera, fast.ai) are NOT in the catalog — refuse.
Briefs DO include apex, name, industry, service_provided, location, ratings. They DON'T include url, phone_primary, email_primary, legal_name, address_full, full platforms — those require an unlock.
not_found / not_in_dataset 404 = not in pro_services. Not charged. Skip.
Unlock is atomic. N apexes either all charge (up to 10×N credits) or none on 402.