The Platform
Eight Modules.
Eight Modules.
One Operating System.
Each module delivers standalone value. Together they form a closed loop from workforce structure to delivery execution — with full governance at every step.
Platform Overview
Three Business Outcomes. Eight Modules.
The modules below map to three strategic outcomes. Each group works standalone — together they close the operating loop.
Measure
Know Your Talent. Precisely.
A real-time, governed picture of every engineer — verified skills, maturity tier, and live availability.
Develop
Grow Engineers. Deliberately.
Structured feedback, continuous achievements, and bias-normalized tiering create a fair, evidence-driven development flywheel.
Deploy
Staff Projects. Confidently.
AI-assisted staffing against verified supply, pre-start delivery risk scoring, and board-ready analytics dashboards.
Talent Roster
The single system of record for every person in your engineering organization.
The Talent Roster is the foundation of the entire platform. Every downstream module — Skills, Feedback, Tiering, Engagements — depends on a clean, validated, and structured roster. Bulk onboarding, manager hierarchy, employment state, and leaver management are all governed here.
Foundation Module
Talent data quality directly determines the quality of every downstream decision — feedback routing, skill gap analysis, staffing match, and leadership reporting all degrade if the roster is stale or inconsistent.
Comprehensive Person Records
Name, role, team, manager, start date, employment status, skills, tier, profile, and full custom field support — all in one governed record.
Bulk CSV Onboarding
Import your entire roster with client-side validation preview before any data is submitted — no silent import failures, no partial writes.
Manager Hierarchy
Self-referential manager relationships power RBAC scope, reportee calculations, and org chart rendering across the entire platform.
Team Move History
Every team assignment is recorded with timestamps — giving you a full operational trace of each person's movement across the organization.
Role-Based PII Masking
Sensitive fields (birth date, gender, exit reason) are masked by default. Admins and the subject retain full access; other roles see only what they need.
Tags and Capabilities
Flexible tagging and capability assignment enables contextual matching beyond formal role profiles — domain expertise, tooling, geography, clearance level.
Skills Intelligence
Governed skill taxonomy with evidence-backed confidence scoring and gap analysis.
Skills is a trust module. If leaders cannot trust that declared skills are real, the entire staffing match and development planning process loses credibility. Every skill claim is governable — proposed, reviewed, evidence-backed, and scored for confidence.
Governed Taxonomy
All skills live in an admin-managed library. Engineers propose skills; team leads or admins approve — preventing taxonomy sprawl and inconsistent naming.
Evidence Types
Attach courses, certifications, and platform badges as evidence. Expired credentials are flagged automatically — no manual audit required.
Confidence Scoring
Self-declared skills score "none". Skills with active certifications score "high". Leaders see which claims are trustworthy at a glance before staffing decisions.
Profile Gap Analysis
Compare each person's skills against their profile's required skill set. Gap severity is calculated per skill: none / partial / missing — with action-ready insights.
Standardized Level Scale
A 5-point level scale (Beginner → Expert) replaces free-text entries, making cross-person gap comparisons meaningful, comparable, and filterable.
Profile-Skill Mapping
Define required skills and minimum required level per role profile — the benchmark against which gap analysis is measured and staffing decisions are made.
Feedback Cycles
Structured multi-pillar reviews with profile-weighted scoring and cycle governance.
Feedback is the signal engine. The quality of tiering, development conversations, and promotion decisions all depend on structured, comparable review signal — not free-text paragraphs or single-score ratings that can't be aggregated.
Six Evaluation Pillars
Delivery Quality, Ownership, Communication, Technical Depth, Architecture Thinking, and Leadership Potential — assessed per cycle with configurable weights per profile.
Profile-Weighted Scoring
Pillar weights are configurable per role profile within a cycle. A Principal Engineer's Technical Depth weighs more than a Graduate's — scoring reflects real role expectations.
Draft and Submit Lifecycle
Reviews are saved as drafts and explicitly submitted. After cycle close, all submissions become immutable — preserving a clean, contestable historical record.
Development Readiness Signal
Each review surfaces a readiness signal (promote / hold / at-risk) as input to the tiering engine — not as an automated decision, but as structured evidence.
Controlled Employee Visibility
Admins control when and whether feedback is visible to the subject — with a per-cycle "visible to employee" flag that can be toggled as your process matures.
Cycle State Machine
Draft → Open → Closed. State transitions are governed — no partial or accidental closures. Full audit trail on every state change, every submission, every edit.
Performance Development
Fiscal-year goals, achievement tracking, and formal reviews — evidence-built, not memory-built.
The strongest selling point of this module: every achievement is formally recorded throughout the year. The year-end review is built from continuous evidence — not a last-minute documentation marathon where managers and engineers try to recall what happened nine months ago.
Goal Templates and Assignments
Reusable goal templates assigned to individuals with fiscal-year scope. Templates separate the reusable structure from person-specific targets and commitments.
Achievement Submission Workflow
Engineers submit achievements as formal evidence records throughout the year — not just at review time. Every achievement is timestamped, categorized, and visible to the reviewing manager.
1-on-1 Meeting Confirmation
Manager confirms 1-on-1 meetings have occurred. A lightweight discipline mechanism that keeps development conversations on track and accountable.
Fiscal Year Scoping
Goals, achievements, and reviews are anchored to a configured fiscal year — aligned with how organizations actually plan, budget, and measure performance.
Controlled Review Visibility
Formal performance reviews have a "visible to employee" flag — admins control when the summary is surfaced to the subject, fitting your HR timeline.
Feedback Cycle Complement
Feedback Cycles provide periodic calibration signal; Performance Development provides the longer-horizon goal and development arc. They are complementary, not competing systems.
Engagements & Staffing
Define work demand as a "Bill of Talent Materials" before staffing begins.
Engagements is the most commercially differentiated module. It turns The Talent Factory from a people record system into a supply-vs-demand operating platform. Demand is formally defined before staffing begins — creating the data discipline most engineering organizations lack entirely.
The Bill of Talent Materials
Just as a manufacturing order defines the exact components required, an Engagement Blueprint defines the exact number of engineers by profile, tier, and skill required for a project. Staffing becomes a data operation, not a conversation.
Blueprint Demand Modeling
Define demand by profile rows: "2× Senior Data Engineers at Tier 2+, with Python and MLflow skills". Structured, comparable across projects, and fully queryable against supply.
AI-Assisted Talent Matching
The staffing engine scores candidates against blueprint requirements using profile match, tier, skills confidence weighting, and current availability — automatically surfacing best fits.
Delivery Risk Scoring
If required talent is unavailable, the system outputs a delivery risk score and surfaces whether an internal promotion review or external hire is the faster remediation path.
Engagement Lifecycle
Draft → Under Review → Approved → Active → On Hold → Closed. Each state transition is governed with business rules — no accidental moves between states.
Revision History
Every blueprint change is tracked. Approval drift is surfaced — showing exactly what changed after an engagement was approved, and by whom.
Configurable Staffing Weights
Customize the matching algorithm: profile weight (30%), tier (25%), skills (30%), capacity (15%). Adjust tier adjacency tolerance and skill level tolerance per organization.
Tiering Engine
Maturity interpretation with governance, bias detection, and full explainability.
Tiering converts feedback signal into a governed workforce maturity view. It is the interpretation layer between signal and decision — sitting between the Feedback module and Engagements demand planning, giving leaders a consistent, bias-corrected view of their talent pool.
Talent Score Formula
Weighted pillar scores from the most recent three closed feedback cycles are computed into a single Talent Score. The formula is fully documented and explainable to every reviewer.
Suggested vs. Final Tier
The system suggests a tier based on the computed score. A human — tenant admin or designated talent council — makes the final classification. Never an automated promotion.
Calibration Index: Bias Detection
Detects grader bias statistically. If a lead's team shows unnatural score clustering, the system flags a Calibration Alert and normalizes scores (Z-score method) to protect engineers from unfair graders.
Admin Override with Rationale
Admins can override suggested tiers with a written rationale. The override and the rationale are stored permanently in the audit trail — no silent changes to career records.
Tier Distribution View
Leadership can see the tier distribution across teams, profiles, and the whole organization — enabling talent density planning and identifying where development investment is needed.
Engagement Demand Integration
Blueprint rows specify tier demand. The tiering engine's output feeds directly into engagement staffing match quality — closing the supply-demand loop automatically.
Analytics & Governance
Leadership dashboards, governance reporting, and complete audit trails across all modules.
Analytics is the visibility layer for every executive and HR leader who needs to answer: "How healthy is my talent supply? Are my review processes producing signal or noise? Where are the critical skill gaps before next quarter's project starts?"
Talent Analytics
Headcount by team, profile mix, tier distribution, attrition risk indicators, and workforce growth trends — real-time and slice-able by team or org unit.
Feedback Analytics
Cycle completion rates, score distributions per pillar, outlier reviewers, and calibration health across cycles — revealing where review process quality needs improvement.
Skills Analytics
Skill coverage by team and profile, gap severity distribution, expiring evidence alerts, and skill confidence health — enabling proactive upskilling investment.
Workforce Utilization
Bench visibility, overallocation alerts, and engagement assignment distribution — giving delivery leaders a real-time operational picture of their talent deployment.
Governance Dashboard
Every sensitive action — tier overrides, admin changes, audit log access, role assignments — is captured and queryable. Full compliance reporting for HR and legal teams.
Module Enablement Controls
Enable or disable each module per tenant. Start with Feedback only. Add Skills and Engagements when your organization is ready — phased adoption without configuration overhead.
Leave Management
Structured leave requests, manager approvals, balance tracking, and availability-aware staffing integration.
The Leave module closes the loop between workforce capacity and project delivery. When engineers submit leave, managers approve, and availability status updates automatically — feeding directly into the Engagements staffing engine to prevent over-assignment and delivery risk during planned absence periods.
Availability-Aware Staffing
Leave data is live in the staffing engine. When an engineer has approved leave overlapping a project start date, they are excluded from staffing recommendations automatically — no manual cross-checking, no silent over-commitment.
Leave Request Workflow
Engineers submit leave requests with type, date range, and optional notes. Managers approve or reject with a comment. The full request-to-decision lifecycle is governed and timestamped.
Leave Balance Tracking
Configurable leave types — annual, sick, personal, unpaid — with per-engineer balance management. Balances update in real-time as requests are approved, giving managers an accurate picture of remaining entitlement.
Engagements Integration
Active and upcoming approved leave is automatically factored into the staffing engine's availability calculation. Engineers on leave cannot be over-staffed during their absence without an explicit override and a documented reason.
Team Leave Calendar
Managers see a unified view of their team's current and upcoming leave. Capacity gaps surface before they become delivery problems — enabling proactive re-planning rather than reactive scrambling.
Leave History & Audit Trail
Every leave request, approval, rejection, and modification is recorded with the actor, timestamp, and comment. Full audit-ready history per engineer — compliant with HR record-keeping requirements.
Admin Policy Configuration
Admins configure leave types, initial balances, accrual rules, and approval workflows per organization. Leave policies are applied consistently across all teams without per-manager customization risk.
Platform Intelligence
7 AI Agents.
Embedded at the Right Moment.
Not a bolt-on chatbot. Each agent is context-aware, purpose-built, and embedded at exactly the moment in the workflow where it creates value — with enterprise privacy controls your security team can approve.
Works with
GPT-4o
GPT-4o mini
Claude Sonnet
Claude Haiku
Gemini
Azure OpenAI
Groq
Ollama
Bring Your Own Key
Feedback Cycles
Feedback Review Assistant
Multi-turn chat embedded inside the review form. Helps managers write structured, evidence-based scores and narratives in real-time — while they draft, not after.
Talent Profiles
Talent Insight Assistant
Chat assistant on the talent profile page. Surfaces the story behind an engineer's career trajectory, tiering history, strengths, and development arc across multiple cycles.
Org-wide · Admin only
Workforce Intelligence
Ask natural language questions about your entire organization: "Who has Python at Tier 2 and is available next month?" or "Where are our critical skill gaps before Q3 projects start?" Answers from live data.
Feedback Cycles
Bias Detection Engine
Analyzes draft scores against observations and prior cycles before submission. Flags leniency, severity, halo effect, recency bias, and similarity bias with a structured 500-word report and remediation advice.
Feedback Cycles
Review Writing Coach
Rewrites draft feedback to be professional, evidence-based, and impact-focused — preserving every factual detail while improving clarity, tone, and usefulness for the engineer receiving it.
Skills Library
Skills Advisor
Analyzes your org's mission, teams, profiles, and existing taxonomy to generate 15 contextually tailored skill recommendations — with rationale and category. Keeps your skills library relevant as your org evolves.
Role Profiles
Profile Builder
Recommends which skills from your taxonomy should be required for each role profile — including the minimum proficiency level for each skill and the rationale behind every mapping decision.
Name Pseudonymization
Real names are replaced with tokens (PERSON_001) before any data reaches the model. Names are restored in the response. Your engineers' identities are never sent to an LLM provider.
Full AI Audit Trail
Every AI request is logged: actor, timestamp, agent type, token counts. Admins see a full observability dashboard with per-module usage, top users, and daily activity — no black box.
DPA Gate & Custom Prompts
AI cannot be enabled without a DPA acknowledgment from the tenant admin. Per-module system prompts and data context flags are configurable — granular control over what each agent is allowed to see and say.
Bring Your Own Key
Connect your own API key for OpenAI, Anthropic Claude, Google Gemini, Azure OpenAI, Groq, or Ollama. Keys are encrypted at rest with AES-128. You own the provider relationship and the data policy.
On the Roadmap
Integrations Coming
The Talent Factory is designed as a standalone operating platform — complete without any integration. Native connectors are on the roadmap to push data where your teams already work.
HRIS
Sync headcount and org structure from BambooHR, Workday, or your HR system of record
Roadmap
GitHub / GitLab
Enrich skill profiles with delivery signals from commit activity and code review patterns
Roadmap
Jira / Linear
Pull project demand and sprint assignments to keep engagement records in sync automatically
Roadmap
Slack
Deliver cycle reminders, staffing alerts, and promotion signals where your managers already live
Roadmap
Want a specific integration prioritized? Let us know →