You’re not brief on instruments. Or fashions. Or frameworks.
What you’re brief on is a principled means to make use of them — at scale.
Constructing efficient generative AI workflows, particularly agentic ones, means navigating a combinatorial explosion of selections.
Each new retriever, immediate technique, textual content splitter, embedding mannequin, or synthesizing LLM multiplies the house of attainable workflows, leading to a search house with over 10²³ attainable configurations.
Trial-and-error doesn’t scale. And model-level benchmarks don’t mirror how elements behave when stitched into full techniques.
That’s why we constructed syftr — an open supply framework for mechanically figuring out Pareto-optimal workflows throughout accuracy, price, and latency constraints.
The complexity behind generative AI workflows
For instance how shortly complexity compounds, contemplate even a comparatively easy RAG pipeline just like the one proven in Determine 1.
Every element—retriever, immediate technique, embedding mannequin, textual content splitter, synthesizing LLM—requires cautious choice and tuning. And past these selections, there’s an increasing panorama of end-to-end workflow methods, from single-agent workflows like ReAct and LATS to multi-agent workflows like CaptainAgent and Magentic-One.
What’s lacking is a scalable, principled solution to discover this configuration house.
That’s the place syftr is available in.
Its open supply framework makes use of multi-objective Bayesian Optimization to effectively seek for Pareto-optimal RAG workflows, balancing price, accuracy, and latency throughout configurations that might be unattainable to check manually.
Benchmarking Pareto-optimal workflows with syftr
As soon as syftr is utilized to a workflow configuration house, it surfaces candidate pipelines that obtain robust tradeoffs throughout key efficiency metrics.
The instance beneath reveals syftr’s output on the CRAG (Complete RAG) Sports activities benchmark, highlighting workflows that preserve excessive accuracy whereas considerably decreasing price.
Whereas Determine 2 reveals what syftr can ship, it’s equally essential to know how these outcomes are achieved.
On the core of syftr is a multi-objective search course of designed to effectively navigate huge workflow configuration areas. The framework prioritizes each efficiency and computational effectivity – important necessities for real-world experimentation at scale.
Since evaluating each workflow on this house isn’t possible, we sometimes consider round 500 workflows per run.
To make this course of much more environment friendly, syftr features a novel early stopping mechanism — Pareto Pruner — which halts analysis of workflows which might be unlikely to enhance the Pareto frontier. This considerably reduces computational price and search time whereas preserving outcome high quality.
Why present benchmarks aren’t sufficient
Whereas mannequin benchmarks, like MMLU, LiveBench, Chatbot Arena, and the Berkeley Function-Calling Leaderboard, have superior our understanding of remoted mannequin capabilities, basis fashions hardly ever function alone in real-world manufacturing environments.
As an alternative, they’re sometimes one element — albeit a necessary one — inside bigger, refined AI techniques.
Measuring intrinsic mannequin efficiency is important, nevertheless it leaves open important system-level questions:
- How do you assemble a workflow that meets task-specific objectives for accuracy, latency, and value?
- Which fashions do you have to use—and wherein components of the pipeline?
syftr addresses this hole by enabling automated, multi-objective analysis throughout complete workflows.
It captures nuanced tradeoffs that emerge solely when elements work together inside a broader pipeline, and systematically explores configuration areas which might be in any other case impractical to judge manually.
syftr is the primary open-source framework particularly designed to mechanically determine Pareto-optimal generative AI workflows that steadiness a number of competing goals concurrently — not simply accuracy, however latency and value as effectively.
It attracts inspiration from current analysis, together with:
- AutoRAG, which focuses solely on optimizing for accuracy
- Kapoor et al. ‘s work, AI Agents That Matter, which emphasizes cost-controlled analysis to forestall incentivizing overly expensive, leaderboard-focused brokers. This precept serves as certainly one of our core analysis inspirations.
Importantly, syftr can also be orthogonal to LLM-as-optimizer frameworks like Trace and TextGrad, and generic stream optimizers like DSPy. Such frameworks may be mixed with syftr to additional optimize prompts in workflows.
In early experiments, syftr first recognized Pareto-optimal workflows on the CRAG Sports activities benchmark.
We then utilized Hint to optimize prompts throughout all of these configurations — taking a two-stage strategy: multi-objective workflow search adopted by fine-grained immediate tuning.
The outcome: notable accuracy enhancements, particularly in low-cost workflows that originally exhibited decrease accuracy (these clustered within the lower-left of the Pareto frontier). These beneficial properties counsel that post-hoc immediate optimization can meaningfully enhance efficiency, even in extremely cost-constrained settings.
This two-stage strategy — first multi-objective configuration search, then immediate refinement — highlights the advantages of mixing syftr with specialised downstream instruments, enabling modular and versatile workflow optimization methods.
Constructing and lengthening syftr’s search house
Syftr cleanly separates the workflow search house from the underlying optimization algorithm. This modular design allows customers to simply prolong or customise the house, including or eradicating flows, fashions, and elements by modifying configuration recordsdata.
The default implementation makes use of Multi-Objective Tree-of-Parzen-Estimators (MOTPE), however syftr helps swapping in different optimization methods.
Contributions of latest flows, modules, or algorithms are welcomed through pull request at github.com/datarobot/syftr.
Constructed on the shoulders of open supply
syftr builds on a lot of highly effective open supply libraries and frameworks:
- Ray for distributing and scaling search over massive clusters of CPUs and GPUs
- Ray Serve for autoscaling mannequin internet hosting
- Optuna for its versatile define-by-run interface (just like PyTorch’s keen execution) and assist for state-of-the-art multi-objective optimization algorithms
- LlamaIndex for constructing refined agentic and non-agentic RAG workflows
- HuggingFace Datasets for quick, collaborative, and uniform dataset interface
- Trace for optimizing textual elements inside workflows, resembling prompts
syftr is framework-agnostic: workflows may be constructed utilizing any orchestration library or modeling stack. This flexibility permits customers to increase or adapt syftr to suit all kinds of tooling preferences.
Case research: syftr on CRAG Sports activities
Benchmark setup
The CRAG benchmark dataset was launched by Meta for the KDD Cup 2024 and contains three duties:
- Activity 1: Retrieval summarization
- Activity 2: Information graph and net retrieval
- Activity 3: Finish-to-end RAG
syftr was evaluated on Activity 3 (CRAG3), which incorporates 4,400 QA pairs spanning a variety of matters. The official benchmark performs RAG over 50 webpages retrieved for every query.
To extend problem, we mixed all webpages throughout all questions right into a single corpus, making a extra sensible, difficult retrieval setting.
Notice: Amazon Q pricing makes use of a per-user/month pricing mannequin, which differs from the per-query token-based price estimates used for syftr workflows.
Key observations and insights
Throughout datasets, syftr persistently surfaces significant optimization patterns:
- Non-agentic workflows dominate the Pareto frontier. They’re quicker and cheaper, main the optimizer to favor these configurations extra steadily than agentic ones.
- GPT-4o-mini steadily seems in Pareto-optimal flows, suggesting it provides a powerful steadiness of high quality and value as a synthesizing LLM.
- Reasoning fashions like o3-mini carry out effectively on quantitative duties (e.g., FinanceBench, InfiniteBench), possible as a result of their multi-hop reasoning capabilities.
- Pareto frontiers finally flatten after an preliminary rise, with diminishing returns in accuracy relative to steep price will increase, underscoring the necessity for instruments like syftr that assist pinpoint environment friendly working factors.
We routinely discover that the workflow on the knee level of the Pareto frontier loses only a few share factors in accuracy in comparison with essentially the most correct setup — whereas being 10x cheaper.
syftr makes it straightforward to seek out that candy spot.
Price of working syftr
In our experiments, we allotted a price range of ~500 workflow evaluations per job. Though actual prices differ based mostly on the dataset and search house complexity, we persistently recognized robust Pareto frontiers with a one-time search price of roughly $500 per use case.
We count on this price to lower as extra environment friendly search algorithms and house definitions are developed.
Importantly, this preliminary funding is minimal relative to the long-term beneficial properties from deploying optimized workflows, whether or not via diminished compute utilization, improved accuracy, or higher consumer expertise in high-traffic techniques.
For detailed outcomes throughout six benchmark duties, together with datasets past CRAG, check with the full syftr paper.
Getting began and contributing
To get began with syftr, clone or fork the repository on GitHub. Benchmark datasets can be found on HuggingFace, and syftr additionally helps user-defined datasets for customized experimentation.
The present search house contains:
- 9 proprietary LLMs
- 11 embedding fashions
- 4 normal immediate methods
- 3 retrievers
- 4 textual content splitters (with parameter configurations)
- 4 agentic RAG flows and 1 non-agentic RAG stream, every with related hierarchical hyperparameters
New elements, resembling fashions, flows, or search modules, may be added or modified through configuration recordsdata. Detailed walkthroughs can be found to assist customization.
syftr is developed absolutely within the open. We welcome contributions through pull requests, function proposals, and benchmark stories. We’re significantly fascinated by concepts that advance the analysis course or enhance the framework’s extensibility.
What’s forward for syftr
syftr continues to be evolving, with a number of energetic areas of analysis designed to increase its capabilities and sensible impression:
- Meta-learning
At present, every search is carried out from scratch. We’re exploring meta-learning methods that leverage prior runs throughout related duties to speed up and information future searches. - Multi-agent workflow analysis
Whereas multi-agent techniques are gaining traction, they introduce extra complexity and value. We’re investigating how these workflows examine to single-agent and non-agentic pipelines, and when their tradeoffs are justified. - Composability with immediate optimization frameworks
syftr is complementary to instruments like DSPy, Hint, and TextGrad, which optimize textual elements inside workflows. We’re exploring methods to extra deeply combine these techniques to collectively optimize construction and language. - Extra agentic duties
We began with question-answer duties, a important manufacturing use case for brokers. Subsequent, we plan to quickly increase syftr’s job repertoire to code technology, information evaluation, and interpretation. We additionally invite the neighborhood to counsel extra duties for syftr to prioritize.
As these efforts progress, we intention to increase syftr’s worth as a analysis instrument, a benchmarking framework, and a sensible assistant for system-level generative AI design.
Should you’re working on this house, we welcome your suggestions, concepts, and contributions.
Strive the code, learn the analysis
To discover syftr additional, try the GitHub repository or read the full paper on ArXiv for particulars on methodology and outcomes.
Syftr has been accepted to seem on the International Conference on Automated Machine Learning (AutoML) in September, 2025 in New York Metropolis.
We stay up for seeing what you construct and discovering what’s subsequent, collectively.
