HackerRank Orchestrate - Support Triage Agent Submission#42
Open
gitsofaryan wants to merge 10 commits into
Open
HackerRank Orchestrate - Support Triage Agent Submission#42gitsofaryan wants to merge 10 commits into
gitsofaryan wants to merge 10 commits into
Conversation
…nd justifications
…and design rationale
Author
|
@sp2hari @anxkhn-hacker @shlokashah I was unable to submit on time due to a 2-day power outage in my village with no electricity access. I have now completed all required deliverables and respectfully request consideration for this submission. Thank you for understanding. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
HackerRank Orchestrate Support Triage Agent
What I Built
A terminal based support ticket agent that intelligently routes tickets across three platforms: HackerRank, Claude, and Visa.
orch.mp4
The system is designed with a strong emphasis on reliability, traceability, and safety. Every decision made by the agent is explainable and grounded in the provided corpus.
How It Works
The agent follows a structured five stage pipeline:
Safety Gate
A multi layer filtering system that detects fraud signals, harmful intent, and unsupported queries before any downstream processing. This ensures risky or ambiguous tickets are escalated early.
Classification
Maps each ticket to the correct company and product area using controlled keyword matching. The focus is on deterministic routing rather than probabilistic guesses.
Hybrid Retrieval
Combines semantic retrieval using FAISS with keyword retrieval using BM25. Results are merged using Reciprocal Rank Fusion to improve relevance and consistency.
LLM Response
Responses are generated through OpenRouter using Claude. If sufficient grounding is not available, the system falls back to extractive responses directly from the corpus. This guarantees zero hallucination.
CSV Output
Writes structured predictions in a spec compliant format with clear justifications for every decision.
Results
Key Features
Deterministic execution with reproducible outputs
Strict corpus grounded responses
Safety first design with gating before LLM usage
Efficient runtime with caching that reduces execution to under five seconds after initial run
Comprehensive documentation including setup and troubleshooting
Submission Files
code/
Modular implementation with clear separation of concerns across 8 modules and a detailed README
support_tickets/output.csv
Predictions for all 29 tickets with reasoning
log.txt
Complete execution trace for debugging and verification
Deliverables
Note
I was unable to submit within the original deadline due to a two day power outage in my village which made it impossible to access a working system. The project has now been fully completed and validated.
I would sincerely appreciate your consideration for this submission.
Thank you for your time.