Unlock Actionable Insights from 360-Degree Feedback with AI

Most companies collect 360 feedback via Google Forms and then manually process hundreds of text comments. The result is lost signal and subjective conclusions. We developed an AI solution that processes up to 15,000 responses in 4 hours, extracts aspects and topics, and builds personalized reports without distortion.

Problem 1: Subjectivity and Scale

When processing 200 questionnaires manually, each analyst inevitably introduces their own biases. Two different HR specialists can interpret the same comment differently: "Ivan runs meetings well but manages timing poorly." The AI system applies a unified aspect-based sentiment analysis (ABSA) model that evaluates each aspect (communication, time management) by the same criteria. We use a fine-tuned BERT on a corpus of HR feedback, achieving an F1-score > 0.87 on the test set (see BERT).

Problem 2: Anonymity Leaks

If a team has only 3–4 reviewers, the stylistic features of a comment reveal the author. Standard pseudonymization methods don't help. We had to implement k-anonymity: the system generalizes or excludes comments with unique stylistic markers from individual reports. Additionally, at the architecture level, we disable author identification: the LLM does not receive reviewer metadata.

The AI System Processes Text Responses Through Three Stages

Each comment goes through three stages:

1. Aspect-based sentiment analysis (ABSA). Determine sentiment per competency aspect, not overall. Use fine-tuned BERT or GPT-4o with structured output.
2. Topic clustering — BERTopic with sentence-transformers embeddings. Clusters are updated each cycle, allowing tracking of issue dynamics.
3. Comparison with self-assessment — compute the discrepancy between self-assessment and peer ratings. If the gap > 1.5 σ, the system flags it for discussion with the manager.

Why AI 360-Degree Feedback Is More Effective?

AI processes 15,000 responses in 4 hours — 30x faster than two HR analysts over 3 weeks. Beyond speed, the model evaluates every statement by consistent criteria, eliminating subjectivity. We guarantee all text responses go through the same pipeline without human bias.

AI Anonymization Using k-Anonymity

The system uses the k-anonymity technique: if a comment is too unique stylistically, it is either generalized (replacing specific words with generic ones) or excluded from the individual report. This ensures that even with a small number of reviewers, the author remains unknown. Additionally, we train the LLM to ignore reviewer metadata — prompt engineering with zero author context.

Practical Case from Our Practice

Our client is a fintech company with 350 employees, a semi-annual 360-feedback cycle. Before implementation: questionnaires in Typeform, manual processing by two HR analysts over 3 weeks, a 5-page report per employee — generic text without specifics. After: collection via custom interface, pipeline using GPT-4o + BERTopic, report generation and dashboard. Cycle time: 4 hours. HR moved from routine to analysis. 78% of employees noted that feedback became specific and actionable (NPS survey).

How to Implement an AI 360-Degree Feedback System: Step-by-Step Guide

Follow these steps:

1. Analyze requirements and design architecture.
2. Integrate with feedback collection tools (Typeform, Google Forms, or internal API).
3. Develop NLP pipeline with analysis, anonymization, and clustering.
4. Build dashboard and reports (Metabase/Grafana or React UI).
5. Test and launch with a pilot group.
6. Train HR team and provide support.

Comparison of Traditional vs. AI Approach

What’s Included in the Work

• Analysis and design: requirements gathering, role profile configuration, model selection.
• Integration: connection to existing collection tools (Typeform, Google Forms, internal API).
• Pipeline development: NLP module, anonymization mechanism, report generation.
• Dashboard: Metabase, Grafana, or custom React UI with filters.
• Training: workshop for HR team on report interpretation.
• Documentation: technical documentation and user guide.
• Support: 3 months post-release support.

Timelines and Cost

Timelines depend on complexity: integration with an existing tool — 3-4 weeks; full system with dashboard — 2-3 months. Implementation costs range from $15,000 to $50,000, with significant ROI through automation. Our company has 10+ years of experience in NLP and has delivered over 50 projects in AI HR automation. Contact us for a consultation and estimate for your task. Request a demo on your data.

Explainable ML: SHAP, LIME, Integrated Gradients, and EU AI Act Requirements

Imagine: a credit scoring model rejects an application. The client demands an explanation, the compliance officer wants detailed documentation. Without built-in explainability (XAI) methods, compliance with modern regulatory requirements is impossible. Our experience includes over 50 projects implementing SHAP, LIME, and Integrated Gradients in production. We guarantee your AI solution becomes transparent, interpretable, and passes audits on the first attempt. Average time to implement basic explanations is 2-4 weeks; full compliance solution takes 6-14 weeks. Contact us for a preliminary assessment of your project.

Why is AI explainability critical for business and compliance?

Explainability is not one task but three distinct requirements. Global explainability shows how the model works overall: which features matter and how they affect predictions on average. Tools: SHAP summary plots, partial dependence plots (PDP), permutation importance. Local explainability explains a specific prediction: why this credit was rejected, which pixels led to a 'cat' classification. Tools: SHAP waterfall, LIME, Integrated Gradients. Contrastive/counterfactual explains what needs to change for a different outcome: 'If income were $10k higher, would the credit be approved?' Tools: DiCE (Diverse Counterfactual Explanations), alibi.

How does SHAP help explain tabular models?

SHAP (SHapley Additive exPlanations) is the standard for tabular data. Based on cooperative game theory: each feature gets a contribution to the deviation of the prediction from the dataset mean. Mathematically correct — satisfies efficiency, symmetry, dummy, and additivity properties.

import shap

explainer = shap.TreeExplainer(lgbm_model)
shap_values = explainer.shap_values(X_test)

# Waterfall plot for a single prediction
shap.plots.waterfall(explainer(X_test)[0])

# Summary for the entire sample
shap.summary_plot(shap_values, X_test, feature_names=feature_names)

TreeExplainer is a fast, accurate algorithm for tree-based models (LightGBM, XGBoost, Random Forest, CatBoost). It computes exact SHAP values in O(TLD²), where T is trees, L is leaves, D is depth. On a model with 1000 trees of depth 6 — milliseconds per explanation. LinearExplainer — for linear models (logistic regression, Ridge) — instant analytical solution. KernelExplainer is model-agnostic, works with any model, but slower: O(2^M) samples for M features. In practice, we use nsamples=1000–5000 as an approximation. For neural networks — DeepExplainer or GradientExplainer.

A common mistake: SHAP values for correlated features are distributed evenly between them — mathematically correct but visually confusing. Features income and income_log have similar SHAP, even though only one is used. Solution: remove duplicate features before training.

When is LIME indispensable?

LIME (Local Interpretable Model-Agnostic Explanations) builds a local linear approximation around the explained example. Faster than SHAP for complex neural networks, but unstable: two runs on the same example may yield different explanations. LIME's strength is text explanations. LimeTextExplainer shows which words influenced the classification. For quick debugging of a text classifier — a convenient tool.

from lime.lime_text import LimeTextExplainer
explainer = LimeTextExplainer(class_names=['neg', 'pos'])
exp = explainer.explain_instance(text, classifier.predict_proba, num_features=10)
exp.show_in_notebook()

What does Integrated Gradients offer for neural networks?

For deep learning models (CNN, Transformer), neither SHAP KernelExplainer nor LIME provides satisfactory explanations: both are too slow or inaccurate. Integrated Gradients (IG) is a gradient-based method, theoretically grounded (axioms completeness, sensitivity, implementation invariance). IG computes the integral of gradients along a straight line from a baseline input (zero or average values) to the actual input. The result is an attribution map showing the contribution of each pixel/token.

from captum.attr import IntegratedGradients

ig = IntegratedGradients(model)
attributions = ig.attribute(
    inputs=input_tensor,
    baselines=baseline_tensor,
    target=predicted_class,
    n_steps=300,
)

The captum library from Meta is the standard for PyTorch. It includes IG, GradCAM, SHAP DeepLift, LayerConductance. GradCAM is simpler, faster, but theoretically weaker. It visualizes which areas of an image the CNN looks at. Sufficient for debugging CV models, but not for compliance documentation.

Comparison of Explainability Methods

EU AI Act: What You Need in Practice

The recently enacted EU AI Act (phased implementation) requires for high-risk systems (credit scoring, medical AI, recruitment systems, law enforcement): technical model documentation, logging of all decisions with audit capability, explanation of each individual decision upon user request, risk assessment and mitigation measures, human oversight. Technically, this means: each prediction must be stored with input features, output, timestamp, model version, and pre-computed explanation. SHAP values are computed at inference and saved alongside the prediction.

For LLM systems, requirements are more complex: there is no standard explanation method, attention weights are not reliable attributions. Current practice: logging full context, retrieved chunks in RAG, chain-of-thought reasoning as proxy explanations. We help determine if the system falls under the high-risk category per Annex III of the EU AI Act, develop a technical model passport (architecture, training data, quality metrics, limitations), configure a decision logging system with retention period (minimum 10 years for some categories), integrate explanation mechanisms into the production pipeline, and implement a user decision appeal procedure.

How We Implement Explainability: Step-by-Step Process

1. Audit and Regulatory Assessment — we determine if the system falls under high-risk category (EU AI Act, GDPR Art. 22, industry requirements Basel IV, MDR). 2-5 days.
2. Integration of Explanations into Inference Pipeline — we connect SHAP, LIME, or IG to the existing service. Configure asynchronous computation with caching. 1-2 weeks.
3. UI Development for Explanations — if a client interface is needed (web dashboard, PDF export). 2-4 weeks.
4. Logging and Audit Setup — we store all inputs, outputs, pre-computed explanations, model version, timestamp. 1-2 weeks.
5. Model Card Documentation — per Google's Model Card Toolkit with breakdown by demographics/subgroups. 1 week.
6. Team Training and Support — documentation handover, engineer training, 3-month SLA support.

What the Result Includes

• Technical model documentation (model card) with intended use, evaluation results by subgroups, limitations, ethical considerations.
• Integrated explanation mechanism (SHAP/LIME/IG) in the production pipeline with automatic saving at inference.
• UI for viewing explanations (web interface or API) with export capability.
• Logging system with retention field configured for EU AI Act requirements.
• Instructions for user decision appeals (for client portal).
• Customer team training (2-3 workshops) and support documentation.

Typical Mistakes in XAI Implementation (and How to Avoid Them)

Feedback and Next Steps

If you need to implement explainability under the EU AI Act, obtain a certified solution, or simply assess the current transparency of your model — request a consultation. We are ready to offer an individualized implementation plan considering your stack (PyTorch, TensorFlow, XGBoost, LLM) and regulatory requirements. Contact us for a detailed cost and timeline estimate for your project.