An AI assistant built into
every part of your delivery workflow.
Dhyey, the Effigate AI assistant, generates tasks, plans sprints, detects blockers, analyses capacity, and answers questions about your live project data — governed, tenant-scoped, and opt-in.
Every AI capability built natively into your delivery workflow
Dhyey is not a bolt-on chatbot. It is embedded directly into project management, sprint planning, helpdesk, and reporting — operating against live tenant data with full governance controls.
Task generation
Dhyey creates tasks and subtasks from user story descriptions, with estimation suggestions and acceptance criteria generated automatically.
Sprint planning AI
AI analyses your backlog and team capacity to recommend sprint composition, velocity-aware estimates, and poker planning support.
Blocker detection
Dhyey surfaces tasks at risk of missing deadlines, dependency conflicts, and capacity overloads before they become delivery problems.
Capacity analysis
AI reviews team member workload across sprints and projects, identifies over-allocation, and suggests rebalancing options.
Sprint summaries
Dhyey generates natural-language sprint summaries for stakeholder review, capturing outcomes, blockers, and velocity trends.
Multi-provider support
Connect Dhyey to OpenAI, Anthropic, Google Gemini, or Ollama for self-hosted deployments — configured per tenant with full credential isolation.
From user story to actionable tasks in seconds
Dhyey reads a user story title and description and generates a structured set of tasks and subtasks that a developer can act on immediately. Estimation suggestions, acceptance criteria, and dependency hints are included so teams spend less time on decomposition and more time on delivery.
- Tasks and subtasks generated from user story descriptions with one action
- Acceptance criteria drafted automatically and editable before saving
- Estimation suggestions based on story complexity and historical team velocity
- Dependency hints surface related stories or tasks that may need to complete first
- All generated content is a proposal — a human confirms before any record is written
Dhyey AI
Task decomposition engine
"As a user I can log in via SSO and have my permissions applied automatically"
AI-assisted sprint composition that respects team capacity
Dhyey analyses the backlog, team capacity, and historical velocity to recommend which stories and tasks should enter the next sprint. Planning poker sessions get AI-generated reference points that teams can discuss rather than starting estimates from scratch, keeping planning sessions shorter and more grounded in data.
- Backlog analysis surfaces the highest-priority and highest-risk items for sprint entry
- Capacity-aware recommendations prevent over-committed sprints from the outset
- Velocity-aware estimates use historical sprint data to calibrate story point suggestions
- Poker planning support provides AI reference estimates alongside team voting
- Sprint composition proposals are fully editable before the sprint is confirmed
Sprint Board
Sprint 12 execution lanes
Auth flow
API rate limits
Dashboard widget
Sprint board UI
Epic details page
Login page
DB schema
CI pipeline
Surface delivery risks before they become missed deadlines
Dhyey continuously monitors open work for tasks approaching their deadline, dependency chains that could block progress, and team members who are over-allocated. Risk signals are presented with context — not just a warning, but a summary of why a task is at risk and what options exist to address it.
- Tasks approaching their deadline flagged with remaining time and completion likelihood
- Dependency conflicts identified where a blocking task is behind schedule
- Capacity overloads detected at the individual and sprint level before they cause delays
- Blocker summaries include context — reason, impact, and suggested resolution paths
- Risk signals surface in sprint board, workload view, and the Dhyey assistant panel
SLA Enforcement Engine
Real-time response and resolution timer governance
P-001: Payment API failure
BreachedCritical · 4h response, 8h resolution
8h 32mP-002: Report export timeout
WarningHigh · 8h response, 16h resolution
14h 22mP-003: UI display glitch
On TrackMedium · 24h response, 48h resolution
11h 04mAI-powered workload balancing across sprints and projects
Dhyey reviews each team member's assigned tasks, logged hours, and sprint commitments to identify over-allocation before performance degrades. When imbalances are found, Dhyey suggests rebalancing options — specific tasks that could move to a less-loaded team member — so managers can act with precision rather than guesswork.
- Individual workload reviewed across all active sprints and projects simultaneously
- Over-allocation detected against declared availability, not just task count
- Rebalancing suggestions identify specific tasks that are good candidates for reassignment
- Capacity forecasts project forward to flag upcoming overloads before the sprint begins
- Workload analysis integrates with the timesheet and leave calendar for accuracy
Workload Capacity
Sprint 12 · Week 2 allocation status
AI that enterprise procurement and security teams can approve
Every AI interaction in Effigate is tenant-scoped, opt-in, and configurable. No tenant data is sent to any AI provider until the tenant has explicitly activated the integration. Credentials are stored per-tenant, data categories sent to AI are configurable, and Dhyey never writes data autonomously — every action requires human confirmation.
- AI features are strictly opt-in per tenant — disabled by default at account creation
- No cross-tenant data access — each AI session operates entirely within the tenant boundary
- AI provider credentials stored per-tenant and never shared across tenants
- Data categories included in AI prompts are configurable in Settings → AI & MCP
- Propose-and-confirm model — Dhyey never creates, updates, or deletes records without explicit user approval
Audit Log Stream
Immutable · tenant-scoped · real-time
admin@acme.com · Token "CI Deploy" created — scope: read
owner@acme.com · INV-2026-0041 issued to Globex Ltd
system · Ticket TK-0182 SLA breached — Critical
system · 14 time logs locked (lock window: 3 days)
owner@acme.com · alice@team.com promoted Owner→Admin
From AI configuration to governed delivery assistance in four steps
Dhyey activates within the boundaries you set and assists across planning, delivery monitoring, and reporting — always with a human in the loop.
-
1Step 01
Configure AI provider
A tenant administrator connects an AI provider (OpenAI, Anthropic, Gemini, or Ollama) in Settings → AI & MCP, sets data sharing preferences, and enables Dhyey for the workspace.
- Provider selected
- Credentials stored per-tenant
- Data categories configured
-
2Step 02
AI assists with planning
Dhyey generates task breakdowns from user stories, suggests sprint compositions, and provides poker planning reference estimates. All proposals are reviewed and confirmed by the team before saving.
- Tasks generated from stories
- Sprint composition suggested
- Estimates proposed for review
-
3Step 03
AI monitors delivery
Throughout the sprint, Dhyey monitors open tasks for deadline risk, dependency conflicts, and capacity overloads. Alerts surface in the sprint board and workload view with actionable context.
- Deadline risks flagged
- Blockers identified
- Capacity overloads detected
-
4Step 04
AI summarises outcomes
At sprint close, Dhyey generates a natural-language sprint summary capturing completed work, unresolved blockers, velocity trends, and recommendations for the next cycle — ready for stakeholder review.
- Sprint summary generated
- Velocity trends captured
- Recommendations proposed
Dhyey delivers differently for every role in the delivery team
AI assistance in Effigate is contextual — the right Dhyey capability surfaces for the right person at the right stage of delivery.
Project Managers
Sprint planning suggestions, capacity forecasts, and sprint summaries reduce administrative overhead and improve planning accuracy.
Developers
Task and subtask generation from user stories removes decomposition overhead and ensures every task has clear acceptance criteria before work begins.
Executives
AI-generated sprint summaries and delivery risk reports provide a concise view of project health without requiring manual status gathering.
Team Leads
Blocker detection and workload rebalancing suggestions surface delivery risks early and provide actionable options to resolve them before SLAs or deadlines are missed.
Connect Dhyey to the AI provider that fits your stack and compliance requirements
Effigate supports leading AI providers and self-hosted models so every team — from cloud-first startups to air-gapped enterprise environments — can run Dhyey on their terms.
OpenAI
Connect Dhyey to GPT-4 and o1 models for task generation and planning assistance.
AI providerAnthropic
Use Claude models for reasoning-intensive sprint planning and report summaries.
AI providerGoogle Gemini
Connect Gemini models for multimodal analysis and delivery insights.
AI providerOllama (self-hosted)
Run local models for air-gapped or privacy-sensitive enterprise environments.
Self-hostedSlack
Receive Dhyey-generated sprint summaries and risk alerts in Slack.
NotificationsQuestions about Dhyey and AI governance in Effigate
No. AI features in Effigate are strictly opt-in per tenant. A tenant administrator must explicitly configure an AI provider and enable Dhyey before any AI calls are made. No tenant data is sent to any external AI API until the tenant has chosen to activate the integration and accepted the relevant data handling terms.
AI provider API keys and connection settings are stored per-tenant in the tenant-scoped database. Credentials are never stored in the landlord database and are never shared across tenants. Each tenant configures their own provider, model selection, and any applicable usage limits independently.
Effigate gives tenants control over what data is included in AI requests. By default, only the minimal context required for the requested operation — such as a user story title and description for task generation — is included. Tenants can review and restrict the data categories sent to AI providers in Settings → AI & MCP. No billing data, rate cards, or client credentials are ever included in AI prompts.
No. Dhyey operates on a propose-and-confirm model. All AI-generated content — tasks, subtasks, sprint compositions, risk reports — is presented as a suggestion that a human must explicitly accept or dismiss before any data is written. Dhyey never creates, updates, or deletes records autonomously.
Effigate supports OpenAI (GPT-4 and o1 series), Anthropic (Claude), Google Gemini, and Ollama for self-hosted local models. Ollama support is specifically designed for air-gapped or privacy-sensitive enterprise environments where data must not leave the on-premises network. Additional provider integrations are added on the roadmap as adoption grows.
Let Dhyey handle the planning overhead — your team focuses on delivery.
14-day free trial. No credit card required. Full platform access from day one.
Related features