Shreya Shankar

Shreya Shankar, Rolando Garcia, Joseph M. Hellerstein et al.

Organizations rely on machine learning engineers (MLEs) to operationalize ML, i.e., deploy and maintain ML pipelines in production. The process of operationalizing ML, or MLOps, consists of a continual loop of (i) data collection and labeling, (ii) experimentation to improve ML performance, (iii) evaluation throughout a multi-staged deployment process, and (iv) monitoring of performance drops in production. When considered together, these responsibilities seem staggering -- how does anyone do MLOps, what are the unaddressed challenges, and what are the implications for tool builders? We conducted semi-structured ethnographic interviews with 18 MLEs working across many applications, including chatbots, autonomous vehicles, and finance. Our interviews expose three variables that govern success for a production ML deployment: Velocity, Validation, and Versioning. We summarize common practices for successful ML experimentation, deployment, and sustaining production performance. Finally, we discuss interviewees' pain points and anti-patterns, with implications for tool design.

Shreya Shankar, Bernease Herman, Aditya G. Parameswaran · uw

While most work on evaluating machine learning (ML) models focuses on computing accuracy on batches of data, tracking accuracy alone in a streaming setting (i.e., unbounded, timestamp-ordered datasets) fails to appropriately identify when models are performing unexpectedly. In this position paper, we discuss how the nature of streaming ML problems introduces new real-world challenges (e.g., delayed arrival of labels) and recommend additional metrics to assess streaming ML performance.

Lindsey Linxi Wei, Shreya Shankar, Sepanta Zeighami et al.

One year ago, we open-sourced DocETL, a declarative system for LLM-powered data processing that, as of March 2026, has 3.7K GitHub stars and users across domains (e.g., journalism, law, medicine, policy, finance, and urban planning). In DocETL, users build pipelines by composing operators described in natural language, also known as semantic operators, with an LLM executing each operator's logic. However, due to complexity in the operator or the data it operates on, LLMs often give inaccurate results. To address this challenge, DocETL introduced rewrite directives, or abstract rules that guide LLM agents in rewriting pipelines by decomposing operators or data. For example, decomposing a single filter("is this email sent from an executive and discussing fraud?") into the conjunction of two separate semantic filters may improve accuracy. However, DocETL only optimizes for accuracy, not cost. How do we optimize for both? We present MOAR (Multi-Objective Agentic Rewrites), a new optimizer for DocETL. To target cost optimization, we introduce two new categories of directives and extend all three existing categories with new ones, bringing the total to over 30 directives -- more than doubling what DocETL originally had. Moreover, since operators can interact with each other unpredictably due to LLM behavior, optimizing operators or sub-pipelines individually can yield suboptimal overall plans. Recognizing this, we design a new global search algorithm that explores rewrites in the context of entire pipelines. Since the space of rewrites is infinite -- pipelines can be rewritten in many ways, and each rewritten pipeline can itself be rewritten -- our algorithm adapts a multi-armed bandit framework to prioritize which pipelines to rewrite. Across six workloads, MOAR achieves 27% higher accuracy than ABACUS, the next-best optimizer, while matching its best accuracy at 55% of its cost.

Ruiying Ma, Shreya Shankar, Ruiqi Chen et al.

Users across enterprises increasingly rely on AI agents to query their data through natural language. However, building reliable data agents remains difficult because real-world data is often fragmented across multiple heterogeneous database systems, with inconsistent references and information buried in unstructured text. Existing benchmarks only tackle individual pieces of this problem -- e.g., translating natural-language questions into SQL queries, answering questions over small tables provided in context -- but do not evaluate the full pipeline of integrating, transforming, and analyzing data across multiple database systems. To fill this gap, we present the Data Agent Benchmark (DAB), grounded in a formative study of enterprise data agent workloads across six industries. DAB comprises 54 queries across 12 datasets, 9 domains, and 4 database management systems. On DAB, the best frontier model (Gemini-3-Pro) achieves only 38% pass@1 accuracy. We benchmark five frontier LLMs, analyze their failure modes, and distill takeaways for future data agent development. Our benchmark and experiment code are published at github.com/ucbepic/DataAgentBench.

Aditya G. Parameswaran, Shreya Shankar, Parth Asawa et al.

Large language models (LLMs) are incredibly powerful at comprehending and generating data in the form of text, but are brittle and error-prone. There has been an advent of toolkits and recipes centered around so-called prompt engineering-the process of asking an LLM to do something via a series of prompts. However, for LLM-powered data processing workflows, in particular, optimizing for quality, while keeping cost bounded, is a tedious, manual process. We put forth a vision for declarative prompt engineering. We view LLMs like crowd workers and leverage ideas from the declarative crowdsourcing literature-including leveraging multiple prompting strategies, ensuring internal consistency, and exploring hybrid-LLM-non-LLM approaches-to make prompt engineering a more principled process. Preliminary case studies on sorting, entity resolution, and imputation demonstrate the promise of our approach

Shreya Shankar, Tristan Chambers, Tarak Shah et al.

Analyzing unstructured data has been a persistent challenge in data processing. Large Language Models (LLMs) have shown promise in this regard, leading to recent proposals for declarative frameworks for LLM-powered processing of unstructured data. However, these frameworks focus on reducing cost when executing user-specified operations using LLMs, rather than improving accuracy, executing most operations as-is (in a single LLM call). This is problematic for complex tasks and data, where LLM outputs for user-defined operations are often inaccurate, even with optimized prompts. For example, an LLM may struggle to identify {\em all} instances of specific clauses, like force majeure or indemnification, in lengthy legal documents, requiring decomposition of the data, the task, or both. We present DocETL, a system that optimizes complex document processing pipelines, while accounting for LLM shortcomings. DocETL offers a declarative interface for users to define such pipelines and uses an agent-based approach to automatically optimize them, leveraging novel agent-based rewrites (that we call rewrite directives), as well as an optimization and evaluation framework. We introduce (i) logical rewriting of pipelines, tailored for LLM-based tasks, (ii) an agent-guided plan evaluation mechanism that synthesizes and orchestrates task-specific validation prompts, and (iii) an optimization algorithm that efficiently finds promising plans, considering the latencies of agent-based plan generation and evaluation. Our evaluation on four different unstructured document analysis tasks demonstrates that DocETL finds plans with outputs that are 25 to 80% more accurate than well-engineered baselines, addressing a critical gap in unstructured data analysis. DocETL is open-source at docetl.org, and as of March 2025, has amassed over 1.7k GitHub Stars, with users spanning a variety of domains.

Youran Sun, Sepanta Zeighami, Bhavya Chopra et al.

Semantic operators have increasingly become integrated within data systems to enable processing data using Large Language Models (LLMs). Despite significant recent effort in improving these operators, their accuracy is limited due to a critical flaw in their implementation: lack of holistic data understanding. In existing systems, semantic operators often process each data record independently using an LLM, without considering data context, only leveraging LLM's dataset-agnostic interpretation of the user-provided task. However, natural language is imprecise, so a task can only be accurately performed if it is correctly interpreted in the context of the dataset. For example, for classification and scoring tasks, which are typical semantic map tasks, the standard method of processing each record row by row yields inaccurate results in a wide range of datasets. We propose HoldUp, a new method for semantic data processing with holistic data understanding. HoldUp processes records jointly, leveraging cross-record relationships to correctly interpret the task within the data context. Enabling holistic data understanding, however, is challenging due to what we call LLM data understanding paradox: while large representative data subsets are necessary to provide context, feeding long inputs to LLMs causes quality degradation due to well-known long-context issues. To resolve this paradox, we develop a novel clustering algorithm to identify the latent structure within the dataset through judicious use of LLMs, inspired by bagging. Using this approach as a primitive, we develop novel clustering-based classification and scoring methods to perform these two tasks with high accuracy. Experiments across 15 real-world datasets show that HoldUp consistently outperforms existing solutions, providing up to 33% higher accuracy for classification and 30% higher accuracy for scoring and clustering tasks.

Reya Vir, Shreya Shankar, Harrison Chase et al.

Large language models (LLMs) are increasingly deployed in specialized production data processing pipelines across diverse domains -- such as finance, marketing, and e-commerce. However, when running them in production across many inputs, they often fail to follow instructions or meet developer expectations. To improve reliability in these applications, creating assertions or guardrails for LLM outputs to run alongside the pipelines is essential. Yet, determining the right set of assertions that capture developer requirements for a task is challenging. In this paper, we introduce PROMPTEVALS, a dataset of 2087 LLM pipeline prompts with 12623 corresponding assertion criteria, sourced from developers using our open-source LLM pipeline tools. This dataset is 5x larger than previous collections. Using a hold-out test split of PROMPTEVALS as a benchmark, we evaluated closed- and open-source models in generating relevant assertions. Notably, our fine-tuned Mistral and Llama 3 models outperform GPT-4o by 20.93% on average, offering both reduced latency and improved performance. We believe our dataset can spur further research in LLM reliability, alignment, and prompt engineering.

Shreya Shankar, J. D. Zamfirescu-Pereira, Björn Hartmann et al. · berkeley

Due to the cumbersome nature of human evaluation and limitations of code-based evaluation, Large Language Models (LLMs) are increasingly being used to assist humans in evaluating LLM outputs. Yet LLM-generated evaluators simply inherit all the problems of the LLMs they evaluate, requiring further human validation. We present a mixed-initiative approach to ``validate the validators'' -- aligning LLM-generated evaluation functions (be it prompts or code) with human requirements. Our interface, EvalGen, provides automated assistance to users in generating evaluation criteria and implementing assertions. While generating candidate implementations (Python functions, LLM grader prompts), EvalGen asks humans to grade a subset of LLM outputs; this feedback is used to select implementations that better align with user grades. A qualitative study finds overall support for EvalGen but underscores the subjectivity and iterative process of alignment. In particular, we identify a phenomenon we dub \emph{criteria drift}: users need criteria to grade outputs, but grading outputs helps users define criteria. What is more, some criteria appears \emph{dependent} on the specific LLM outputs observed (rather than independent criteria that can be defined \emph{a priori}), raising serious questions for approaches that assume the independence of evaluation from observation of model outputs. We present our interface and implementation details, a comparison of our algorithm with a baseline approach, and implications for the design of future LLM evaluation assistants.

Shreya Shankar, Yoni Halpern, Eric Breck et al.

Modern machine learning systems such as image classifiers rely heavily on large scale data sets for training. Such data sets are costly to create, thus in practice a small number of freely available, open source data sets are widely used. We suggest that examining the geo-diversity of open data sets is critical before adopting a data set for use cases in the developing world. We analyze two large, publicly available image data sets to assess geo-diversity and find that these data sets appear to exhibit an observable amerocentric and eurocentric representation bias. Further, we analyze classifiers trained on these data sets to assess the impact of these training distributions and find strong differences in the relative performance on images from different locales. These results emphasize the need to ensure geo-representation when constructing data sets for use in the developing world.

Shu Liu, Soujanya Ponnapalli, Shreya Shankar et al.

Large Language Model (LLM) agents, acting on their users' behalf to manipulate and analyze data, are likely to become the dominant workload for data systems in the future. When working with data, agents employ a high-throughput process of exploration and solution formulation for the given task, one we call agentic speculation. The sheer volume and inefficiencies of agentic speculation can pose challenges for present-day data systems. We argue that data systems need to adapt to more natively support agentic workloads. We take advantage of the characteristics of agentic speculation that we identify, i.e., scale, heterogeneity, redundancy, and steerability - to outline a number of new research opportunities for a new agent-first data systems architecture, ranging from new query interfaces, to new query processing techniques, to new agentic memory stores.

Quentin Romero Lauro, Shreya Shankar, Sepanta Zeighami et al.

Retrieval-augmented generation (RAG) pipelines have become the de-facto approach for building AI assistants with access to external, domain-specific knowledge. Given a user query, RAG pipelines typically first retrieve (R) relevant information from external sources, before invoking a Large Language Model (LLM), augmented (A) with this information, to generate (G) responses. Modern RAG pipelines frequently chain multiple retrieval and generation components, in any order. However, developing effective RAG pipelines is challenging because retrieval and generation components are intertwined, making it hard to identify which component(s) cause errors in the eventual output. The parameters with the greatest impact on output quality often require hours of pre-processing after each change, creating prohibitively slow feedback cycles. To address these challenges, we present RAGGY, a developer tool that combines a Python library of composable RAG primitives with an interactive interface for real-time debugging. We contribute the design and implementation of RAGGY, insights into expert debugging patterns through a qualitative study with 12 engineers, and design implications for future RAG tools that better align with developers' natural workflows.

Sepanta Zeighami, Shreya Shankar, Aditya Parameswaran

Large Language Models (LLMs) are being increasingly used as a building block in data systems to process large text datasets. To do so, LLM model providers offer multiple LLMs with different sizes, spanning various cost-quality trade-offs when processing text at scale. Top-of-the-line LLMs (e.g., GPT-4o, Claude Sonnet) operate with high accuracy but are prohibitively expensive when processing many records. To avoid high costs, more affordable but lower quality LLMs (e.g., GPT-4o-mini, Claude Haiku) can be used to process records, but we need to ensure that the overall accuracy does not deviate substantially from that of the top-of-the-line LLMs. The model cascade framework provides a blueprint to manage this trade-off, by using the confidence of LLMs in their output (e.g., log-probabilities) to decide on which records to use the affordable LLM. However, existing solutions following this framework provide only marginal cost savings and weak theoretical guarantees because of poor estimation of the quality of the affordable LLM's outputs. We present BARGAIN, a method that judiciously uses affordable LLMs in data processing to significantly reduce cost while providing strong theoretical guarantees on the solution quality. BARGAIN employs a novel adaptive sampling strategy and statistical estimation procedure that uses data and task characteristics and builds on recent statistical tools to make accurate estimations with tight theoretical guarantees. Variants of BARGAIN can support guarantees on accuracy, precision, or recall of the output. Experimental results across 8 real-world datasets show that BARGAIN reduces cost, on average, by up to 86% more than state-of-the-art, while providing stronger theoretical guarantees on accuracy of output, with similar gains when guaranteeing a desired level of precision or recall.

Sepanta Zeighami, Yiming Lin, Shreya Shankar et al.

With the power of LLMs, we now have the ability to query data that was previously impossible to query, including text, images, and video. However, despite this enormous potential, most present-day data systems that leverage LLMs are reactive, reflecting our community's desire to map LLMs to known abstractions. Most data systems treat LLMs as an opaque black box that operates on user inputs and data as is, optimizing them much like any other approximate, expensive UDFs, in conjunction with other relational operators. Such data systems do as they are told, but fail to understand and leverage what the LLM is being asked to do (i.e. the underlying operations, which may be error-prone), the data the LLM is operating on (e.g., long, complex documents), or what the user really needs. They don't take advantage of the characteristics of the operations and/or the data at hand, or ensure correctness of results when there are imprecisions and ambiguities. We argue that data systems instead need to be proactive: they need to be given more agency -- armed with the power of LLMs -- to understand and rework the user inputs and the data and to make decisions on how the operations and the data should be represented and processed. By allowing the data system to parse, rewrite, and decompose user inputs and data, or to interact with the user in ways that go beyond the standard single-shot query-result paradigm, the data system is able to address user needs more efficiently and effectively. These new capabilities lead to a rich design space where the data system takes more initiative: they are empowered to perform optimization based on the transformation operations, data characteristics, and user intent. We discuss various successful examples of how this framework has been and can be applied in real-world tasks, and present future directions for this ambitious research agenda.

Shreya Shankar, Aditya Parameswaran

Software organizations are increasingly incorporating machine learning (ML) into their product offerings, driving a need for new data management tools. Many of these tools facilitate the initial development of ML applications, but sustaining these applications post-deployment is difficult due to lack of real-time feedback (i.e., labels) for predictions and silent failures that could occur at any component of the ML pipeline (e.g., data distribution shift or anomalous features). We propose a new type of data management system that offers end-to-end observability, or visibility into complex system behavior, for deployed ML pipelines through assisted (1) detection, (2) diagnosis, and (3) reaction to ML-related bugs. We describe new research challenges and suggest preliminary solution ideas in all three aspects. Finally, we introduce an example architecture for a "bolt-on" ML observability system, or one that wraps around existing tools in the stack.

Sumanth Dathathri, Krishnamurthy Dvijotham, Alexey Kurakin et al.

Convex relaxations have emerged as a promising approach for verifying desirable properties of neural networks like robustness to adversarial perturbations. Widely used Linear Programming (LP) relaxations only work well when networks are trained to facilitate verification. This precludes applications that involve verification-agnostic networks, i.e., networks not specially trained for verification. On the other hand, semidefinite programming (SDP) relaxations have successfully be applied to verification-agnostic networks, but do not currently scale beyond small networks due to poor time and space asymptotics. In this work, we propose a first-order dual SDP algorithm that (1) requires memory only linear in the total number of network activations, (2) only requires a fixed number of forward/backward passes through the network per iteration. By exploiting iterative eigenvector methods, we express all solver operations in terms of forward and backward passes through the network, enabling efficient use of hardware like GPUs/TPUs. For two verification-agnostic networks on MNIST and CIFAR-10, we significantly improve L-inf verified robust accuracy from 1% to 88% and 6% to 40% respectively. We also demonstrate tight verification of a quadratic stability specification for the decoder of a variational autoencoder.

Rohit Jammula, Vishnu Rajan Tejus, Shreya Shankar

Deep learning models have the capacity to fundamentally revolutionize medical imaging analysis, and they have particularly interesting applications in computer-aided diagnosis. We attempt to use deep learning neural networks to diagnose funduscopic images, visual representations of the interior of the eye. Recently, a few robust deep learning approaches have performed binary classification to infer the presence of a specific ocular disease, such as glaucoma or diabetic retinopathy. In an effort to broaden the applications of computer-aided ocular disease diagnosis, we propose a unifying model for disease classification: low-cost inference of a fundus image to determine whether it is healthy or diseased. To achieve this, we use transfer learning techniques, which retain the more overarching capabilities of a pre-trained base architecture but can adapt to another dataset. For comparisons, we then develop a custom heuristic equation and evaluation metric ranking system to determine the optimal base architecture and hyperparameters. The Xception base architecture, Adam optimizer, and mean squared error loss function perform best, achieving 90% accuracy, 94% sensitivity, and 86% specificity. For additional ease of use, we contain the model in a web interface whose file chooser can access the local filesystem, allowing for use on any internet-connected device: mobile, PC, or otherwise.

Gamaleldin F. Elsayed, Shreya Shankar, Brian Cheung et al.

Machine learning models are vulnerable to adversarial examples: small changes to images can cause computer vision models to make mistakes such as identifying a school bus as an ostrich. However, it is still an open question whether humans are prone to similar mistakes. Here, we address this question by leveraging recent techniques that transfer adversarial examples from computer vision models with known parameters and architecture to other models with unknown parameters and architecture, and by matching the initial processing of the human visual system. We find that adversarial examples that strongly transfer across computer vision models influence the classifications made by time-limited human observers.