Salesforce

8 production-ready jobs — CRM-to-persona pipelines, deal-stage focus groups, account intelligence, brand voice extraction, lead scoring validation, pipeline content, competitive battle cards, and QBR prep

Salesforce holds your richest buyer data — contacts, opportunities, accounts, notes, cases. These eight jobs pull that data through Mavera’s surfaces (Personas, Focus Groups, Mave Agent, Brand Voices, Generations) so every campaign, battle card, and QBR is grounded in real CRM intelligence.

---

API Reference Card

Detail	Value
Base URL	https://{your-domain}.my.salesforce.com/services/data/v66.0/
Auth	OAuth 2.0 — Web Server flow, JWT Bearer, or Client Credentials
Rate limits	100,000 requests / rolling 24 h (Enterprise); 25 concurrent long-running
Mavera base	https://app.mavera.io/api/v1
Mavera auth	Authorization: Bearer mvra_live_...

All examples use three environment variables: SALESFORCE_INSTANCE (e.g. yourorg.my.salesforce.com), SALESFORCE_ACCESS_TOKEN, and MAVERA_API_KEY. Store them in your secret manager — never commit tokens.

---

Jobs Directory

#	Job	Description
1	CRM-to-Persona Pipeline	Pull contacts via SOQL, group by title/industry, create data-grounded personas
2	Deal-Stage Focus Group	Win/loss analysis via synthetic focus groups mapped to opportunity stages
3	Account Intelligence Brief	AI research briefs from Account, Contact, and Task data
4	Sales Note → Brand Voice	Extract winning language from Closed Won deal notes into a Brand Voice profile
5	Lead Scoring Validation	Validate lead scoring models with synthetic focus group feedback
6	Pipeline-Aware Content Generation	Generate stage-specific nurture content tied to your real pipeline
7	Competitive Displacement Tracker	AI-generated battle cards from competitor mentions in opportunities
8	Customer Success Interview Prep	QBR prep documents from Account health, Cases, and Activities

CRM-to-Persona Pipeline

Pull contacts via SOQL, group by title/industry, create data-grounded personas

Deal-Stage Focus Group

Win/loss analysis via synthetic focus groups mapped to opportunity stages

Account Intelligence Brief

AI research briefs from Account, Contact, and Task data

Sales Note → Brand Voice

Extract winning language from Closed Won deal notes into a Brand Voice profile

Lead Scoring Validation

Validate lead scoring models with synthetic focus group feedback

Pipeline-Aware Content Generation

Generate stage-specific nurture content tied to your real pipeline

Competitive Displacement Tracker

AI-generated battle cards from competitor mentions in opportunities

Customer Success Interview Prep

QBR prep documents from Account health, Cases, and Activities

---

Production Tips

Token refresh — Access tokens expire. Use the OAuth 2.0 refresh token flow or JWT Bearer for server-to-server jobs. Never hardcode tokens.

Rate limit budgeting — Enterprise orgs get 100,000 API calls per rolling 24 hours. Each SOQL query counts as one call. Batch queries and cache responses. Monitor usage at GET /services/data/v66.0/limits.

SOQL pagination — Queries returning more than 2,000 rows include a nextRecordsUrl field. Follow it until done: true:

results = []
url = f"https://{SF}/services/data/v66.0/query?q={soql}"
while url:
    resp = requests.get(url, headers=SF_HEADERS).json()
    results.extend(resp["records"])
    url = resp.get("nextRecordsUrl")
    if url:
        url = f"https://{SF}{url}"

Error handling checklist:

Error	Cause	Fix
401 Unauthorized	Expired or invalid SF token	Refresh via OAuth; check Connected App scopes
403 Forbidden	User lacks object/field permissions	Verify profile permissions; check field-level security
429 Too Many Requests	Rate limit exceeded	Back off exponentially; check /limits endpoint
400 MALFORMED_QUERY	SOQL syntax error	Validate query in Developer Console first
INVALID_FIELD	Custom field API name wrong	Use Describe endpoint to verify field names
Mavera 401	Invalid or expired API key	Rotate key at app.mavera.io/settings
Mavera 422	Malformed request body	Check required fields in API docs

Concurrency — Salesforce allows 25 concurrent long-running requests. For batch jobs across hundreds of accounts, use a semaphore or queue (asyncio.Semaphore(25) in Python, p-limit(25) in Node).

Custom field discovery — Not sure if Competitor__c or Health_Score__c exists in your org? Use the Describe endpoint:

GET /services/data/v66.0/sobjects/Opportunity/describe

---

What’s Next

HubSpot Integration

8 jobs — lifecycle personas, deal replay, meeting analysis

Personas API

Full reference for POST /api/v1/personas

Focus Groups API

Full reference for POST /api/v1/focus-groups

Mave Agent

Full reference for POST /api/v1/mave/chat