StockGro6

Mission Control · Agent Roster

Fifteen agents, each with a mission and a human owner

Autonomy is granted per agent, not per company. An agent only earns a higher ceiling when its eval score holds for two weeks and its escalation rate stays inside the guardrail.

Agents live

12/14

Mean eval score

87

weekly harness, 300 cases each

Human escalations today

2,702

by design, not by failure

Model spend today

₹2.9 L

across every squad

Stoxo Desk

Answers member research questions across 4,100 tracked instruments and cites every source.

act-with-approval
@stoxoResearchlast run 20714d ago
Human owner
Research Ops
Model
reasoning · long-context
Runs today
18,420
Success rate
96.4%
Escalated to a human
142
Minutes saved
9,800
Cost today
₹41.2k

Tools it may call

filings-indexprice-apianalyst-notescompliance-check

91

/100

eval score

7-week eval trend

Success rate96.4%

Compliance Guard

Pre-screens every ad, reel, push and advisory message against SEBI advertisement and RA norms.

act-with-approval
@guardRisklast run 20714d ago
Human owner
Compliance
Model
policy-tuned
Runs today
1,284
Success rate
99.1%
Escalated to a human
26
Minutes saved
3,120
Cost today
₹8.6k

Tools it may call

sebi-rulebookcreative-ocrarchive-writeraudit-ledger

96

/100

eval score

7-week eval trend

Success rate99.1%

Retention Sentinel

Watches 214k paid members for churn signals and runs save-plays before renewal date.

act-with-approval
@keepRevenuelast run 20714d ago
Human owner
Growth
Model
propensity + reasoning
Runs today
7,640
Success rate
88.2%
Escalated to a human
318
Minutes saved
5,400
Cost today
₹22.4k

Tools it may call

subscription-dbusage-eventswhatsapp-sendoffer-engine

84

/100

eval score

7-week eval trend

Success rate88.2%

Voice Concierge

Hindi/English outbound + inbound voice for activation, renewals and KYC nudges.

act-with-approval
@voiceConversationslast run 20714d ago
Human owner
Member Ops
Model
realtime speech
Runs today
11,260
Success rate
82.7%
Escalated to a human
604
Minutes saved
14,200
Cost today
₹63.8k

Tools it may call

telephonykyc-statusbroker-linkcrm-write

79

/100

eval score

7-week eval trend

Success rate82.7%

Support Resolver

Owns tier-1 member support end to end: payments, app issues, plan changes, refunds.

autonomous
@resolveConversationslast run 20714d ago
Human owner
Member Ops
Model
fast chat
Runs today
24,180
Success rate
91.5%
Escalated to a human
1,120
Minutes saved
18,600
Cost today
₹29.4k

Tools it may call

ticket-dbpayments-apirefund-policyapp-logs

88

/100

eval score

7-week eval trend

Success rate91.5%

Activation Pilot

Drives signup → KYC → broker link → first trade, unblocking drop-offs within 9 minutes.

act-with-approval
@activateRevenuelast run 20714d ago
Human owner
Growth
Model
orchestrator
Runs today
9,320
Success rate
86.1%
Escalated to a human
212
Minutes saved
7,300
Cost today
₹18.9k

Tools it may call

kyc-vendorbroker-linkpush-sendvoice-handoff

83

/100

eval score

7-week eval trend

Success rate86.1%

Creative Forge

Generates and variants ad creative, reels scripts and hooks per cohort, in 7 languages.

suggest
@forgeMarketinglast run 20714d ago
Human owner
Marketing
Model
multimodal
Runs today
860
Success rate
93.4%
Escalated to a human
64
Minutes saved
4,100
Cost today
₹16.2k

Tools it may call

brand-kitimage-genhook-librarycompliance-check

87

/100

eval score

7-week eval trend

Success rate93.4%

Media Buyer

Reallocates ₹ across Meta/Google/ASA every 30 minutes against blended CAC-to-LTV.

act-with-approval
@spendMarketinglast run 20714d ago
Human owner
Performance
Model
bandit + reasoning
Runs today
48
Success rate
90.6%
Escalated to a human
9
Minutes saved
2,600
Cost today
₹5.4k

Tools it may call

meta-apigoogle-adsasaltv-model

85

/100

eval score

7-week eval trend

Success rate90.6%

Market Radar

Listens to markets, regulators, app reviews and social chatter; raises what changes the narrative.

suggest
@radarResearchlast run 20714d ago
Human owner
Research Ops
Model
streaming + reasoning
Runs today
2,140
Success rate
94.2%
Escalated to a human
41
Minutes saved
3,900
Cost today
₹11.8k

Tools it may call

news-feedx-listenreddit-listenstore-reviews

89

/100

eval score

7-week eval trend

Success rate94.2%

Advisor Coach

Scores every advisor and counsellor call, then drills the exact weak moment next morning.

suggest
@coachEnablementlast run 20714d ago
Human owner
Enablement
Model
audio + reasoning
Runs today
1,420
Success rate
92.8%
Escalated to a human
18
Minutes saved
2,900
Cost today
₹7.4k

Tools it may call

call-recordingsscorecarddrill-builderlms

86

/100

eval score

7-week eval trend

Success rate92.8%

Creator Scout

Finds, scores and briefs finfluencers; flags anyone whose content would breach advisory norms.

suggest
@scoutMarketinglast run 20714d ago
Human owner
Marketing
Model
multimodal
Runs today
312
Success rate
87.9%
Escalated to a human
27
Minutes saved
1,800
Cost today
₹4.3k

Tools it may call

social-graphbrand-safetyrate-card-dbcontract-gen

81

/100

eval score

7-week eval trend

Success rate87.9%

Stoxo B2B Steward

Runs white-label Stoxo tenants for brokers and banks: quotas, evals, tenant-safe answers.

act-with-approval
@b2bNew Revenuelast run 20714d ago
Human owner
New Revenue
Model
orchestrator
Runs today
5,820
Success rate
97.2%
Escalated to a human
11
Minutes saved
2,400
Cost today
₹21.4k

Tools it may call

tenant-registryusage-meteringsla-monitoreval-harness

92

/100

eval score

7-week eval trend

Success rate97.2%

Academy Tutor

Personal trading-discipline curriculum with certification, feeding straight into paid plans.

autonomous
@tutorEnablementlast run 20714d ago
Human owner
Academy
Model
tutor-tuned
Runs today
15,600
Success rate
94.8%
Escalated to a human
72
Minutes saved
6,100
Cost today
₹19.7k

Tools it may call

curriculumbacktest-sandboxquiz-gencertificates

88

/100

eval score

7-week eval trend

Success rate94.8%

Research Factory

Turns analyst conviction into 40 formats a day — notes, reels, push, WhatsApp, newsletter.

act-with-approval
@factoryResearchlast run 20714d ago
Human owner
Research Ops
Model
long-context
Runs today
1,180
Success rate
95.3%
Escalated to a human
38
Minutes saved
5,200
Cost today
₹14.6k

Tools it may call

analyst-draftsformat-templatescompliance-checkpublisher

90

/100

eval score

7-week eval trend

Success rate95.3%

How an agent earns more autonomy

DeployOne's promotion ladder, applied per agent

  1. 1 · Observe — reads production data, writes no action. Its proposals are scored against what the human actually did.
  2. 2 · Suggest — drafts the action; a human presses the button. Agreement rate must hold above 85% for two weeks.
  3. 3 · Act with approval — executes once a named human clears it. Money and member-facing copy stay here permanently.
  4. 4 · Autonomous — acts alone inside a hard blast radius, with reversal built in and a daily sampled audit.