arXiv subjects

Divya Chukkapalli

Publications and source records attributed to Divya Chukkapalli.

5 recordsLinked to original sources

arXiv2026

DRL: A Deterministic Relational Middleware Layer for Transaction-Safe Enterprise NL2SQL Under Schema-Graph Scaling

Deploying natural-language interfaces over enterprise OLTP catalogs fails at scale because semantic parsers collapse under schema-graph scaling, inflating context beyond stable LLM attention budgets. We present DRL (Deterministic Relational Middleware Layer), a safe pipeline interposing between front-ends and SQL backends. DRL comprises dynamic context pruning, relational AST typing, and transactional safeguard verification (EXPLAIN gating and NULL guards) to bound context and flag operational silent divergence (SDop). We evaluate DRL on PostgreSQL and MySQL, contributing (i) an OLTP schema-graph scaling model, (ii) a 1,000-pair Workload Verification Suite, (iii) baselines B0-B3, and (iv) an enterprise NL2SQL failure taxonomy. On PostgreSQL, schema-linked hints (B1) yield a 76% context reduction over naive full-catalog prompting (B0); DRL's dynamic router (B2) reaches a 92% reduction at pruning p95 = 0.58 ms and middleware p95 = 4.6 ms. GPT-4o, Claude Sonnet 4.5, and Gemini 2.5 Flash achieve 52.9%, 52.8%, and 52.1% execution match under a corrected evaluation harness; SDop flags 89-100% of false-positive EX-passing queries. GPT-4o failures are dominated by semantic/filter errors (254/471), while column hallucination is a minor factor (47/471). Crucially, a single regex defect in our evaluation post-processor silently suppressed accuracy and manufactured a false 4-10% cross-vendor gap that vanished when corrected, showing that benchmark code deserves the same scrutiny as the models it scores. DRL reframes enterprise NL2SQL as systems engineering - context bounding, verification, and plan-aware admission - not a leaderboard exercise.

cs.DB↗

arXiv2026

ESQ-Bench: A Multi-Tier Enterprise Oracle Benchmark for Evaluating NL2SQL Dialect Generalization and Silent Semantic Divergence

State-of-the-art Natural Language to SQL (NL2SQL) models report execution accuracy exceeding 89 percent on established benchmarks such as Spider and BIRD. However, these benchmarks rely on simplified academic schemas and open-source SQL dialects that do not reflect the complexity of enterprise database environments. We introduce ESQ-Bench, an Oracle-first NL2SQL benchmark with systematic complexity tiers and silent-divergence evaluation across three enterprise schema complexity tiers. We constructed and released six populated schemas (465 tables, 164,682 rows, zero empty tables) with identical seed data on Oracle, PostgreSQL, MySQL, and SQL Server, a four-metric evaluation harness (EM, EX, SR, SD), and 550 gold-validated question-query pairs (Tier-1: 95; Tier-2: 228; Tier-3: 227). Schema-linked prompting with GPT-4o shows monotonic execution-match degradation across tiers: 79.8, 60.3, and 57.2 percent EX on executed queries (June 2026), versus 75.6, 80.4, and 95.8 percent on an earlier 142-question pilot slice. EM stays below 7 percent tier-wide; operational silent-divergence reaches 73 to 99 percent among EX-passing queries. Failure analysis shows wrong-result semantics dominate at higher tiers. Claude Sonnet 4.6 with schema-linked prompts reaches 87.4, 74.9, and 68.7 percent EX (executed queries), exceeding GPT-4o schema-linked on every tier. GPT-4o zero-shot EX on executed queries (78.7, 73.5, and 77.8 percent) inverts schema-linked at Tiers 2 to 3 due to lower execution rates and survivor bias in the zero-shot versus schema-linked analysis. Local Llama 3.2 schema-linked reaches only 13.3 percent bank-wide EX (73 out of 550), underscoring the gap between closed API models and open-weight baselines on enterprise Oracle schemas.

cs.AI↗

arXiv2026

Stateful Guardrails for Multi-Turn LLM Systems: A Conversational Risk Accumulation Framework

Most safety guardrails for large language models (LLMs) evaluate each prompt-response pair in isolation, which misses failures that arise only over a dialogue as benign turns compose into harm. We term this Conversational Risk Accumulation (CRA): gradual intent drift, fragmented assembly of prohibited instructions, and sensitivity build-up from repeated disclosures. We propose a session-layer CRA Framework that tracks three trajectory signals: semantic drift from a session anchor, a sensitivity-weighted information accumulation graph over extracted entities, and a compliance-gradient signal capturing increasing willingness to comply. For scoring, we provide (i) an unsupervised convex fusion for attribution and ablations, and (ii) CRA-Net DA, a compact learned trajectory model trained with family-adversarial objectives to reduce length and topic-coverage confounds. To benchmark CRA, we release CRA-Bench v0.1 (1,200 eight-turn sessions across three threat families with topic-matched benign twins), CRA-Bench v0.2 (LLM-paraphrased variants to reduce template artifacts), and an extended 5-family set (2,000 sessions adding persona priming and context stuffing). We introduce a trajectory-native evaluation protocol with session-level splits, mixed-set threshold calibration, Trajectory AUROC, turns-to-detection, calibrated false-positive metrics, bootstrap confidence intervals, leave-one-family-out diagnostic stress tests, and synthetic-to-human transfer checks. Claims focus on within-distribution session scoring on CRA-Bench and human-transfer subsets.

cs.CL↗

arXiv2026

Schema-Aware Localisation (SAL): Live Schema Grounding and Hallucination Validation for Oracle NL2SQL

Large language models can generate fluent SQL from natural language, but on real enterprise Oracle databases they frequently fail at execution time: columns and aliases are hallucinated and dialect-specific syntax is missed, leading to ORA-00904 invalid-identifier errors. In this setting, failures are primarily due to missing schema grounding: the model cannot know which tables and columns actually exist. This paper introduces Schema-Aware Localisation (SAL), a lightweight middleware layer for Oracle NL2SQL that requires no model retraining. SAL queries Oracle's USER_TAB_COLUMNS catalog to build a live schema map, selects a relevant table subset for each question (falling back to the full schema for multi-table queries), and injects this ground-truth context into the LLM prompt. Generated SQL is then checked by the Hallucination Index (Hidx), which validates every alias.column reference against the live catalog, automatically rewrites predictable prefix errors, and otherwise triggers a structured retry with itemised corrections. We evaluate SAL on 500 TPC-H natural language questions executed against a live Oracle Autonomous Database 23c instance using GPT-4o-mini. Without any schema grounding, execution-grounded truth (EGT; executes and matches the reference result set) is 2.2% (12/500). A hand-written static schema hint brings EGT to 62.0%. SAL, with no manual schema curation, achieves 62.6% EGT (96% simple, 95% medium, 40.7% complex) while reducing execution failures from 97.6% to 2.6%.

cs.AI↗

arXiv2026

CommitDistill: A Lightweight Knowledge-Centric Memory Layer for Software Repositories

Software repositories accumulate large amounts of unstructured knowledge in commit messages, pull-request discussions, and issue threads, but developers and AI coding assistants rarely reuse this history effectively. Recent work on typed-memory architectures for LLM agents (MemGPT, generative agents, and the PlugMem module of Yang et al.) argues that agent memory should be distilled, typed knowledge rather than raw interaction text. We adapt that stance to a software repository's own git history under a constrained regime: deterministic, dependency-free, local-only, no embeddings. We present CommitDistill, an open-source Python prototype that mines a local git history into typed knowledge units (Facts, Skills, Patterns) using deterministic regex and surfaces them through a TF-IDF retriever with a calibrated silence threshold (theta = 2.5) that abstains on out-of-distribution queries. The artefact is a trust-instrumented memory substrate: deterministic, no external service, inspectable plain-JSON store, tunable abstention. A case study on five public repositories spanning Python, JavaScript, C, and Java (25,000 commits, 1,167 extracted units) reports useful-precision 0.525 at Cohen's kappa = 0.633 on 40 dual-annotated Python units. The decisive finding is budget-constrained retrieval: at a 256-character per-query budget, CommitDistill reaches 0.750 hit-rate on a 12-query benchmark against BM25's 0.333 and git log --grep's 0.083. On a four-arm paired LLM-as-judge evaluation (n=200 time-travel bug-fixes, two judges) covering control, CommitDistill, a body-budget-matched CD-Hybrid, and BM25, no condition produces a statistically detectable lift over control on the headline mean and CD-Hybrid is indistinguishable from BM25 head-to-head. Extraction over 10,000 commits completes in under 4 seconds on a laptop. Source, annotations, baselines, and a reproducibility script accompany this paper.

cs.SE↗