UCL University College London
f.sarro@ucl.ac.uk | ORCID: 0000-0002-9146-442X
Prof. Federica Sarro
Prof. Federica Sarro
Professor of Software Engineering
Chair, IEEE CS TCSE Lero Parnas Fellow & IEEE CS Rising Star ACM Distinguished Speaker

Professor Federica Sarro

Full Professor of Software Engineering • University College London

Departmental Leadership
Head of Software Systems Engineering (SSE) Group
UCL Computer Science
Research Laboratory
Founder & Director of SOLAR Group
Software Optimisation, Learning & Analytics
Global Community
Chair of IEEE CS TCSE
Technical Community on Software Engineering

Professor Federica Sarro leads pioneering research in Search-Based Software Engineering (SBSE), Software Fairness & Responsible AI, LLMs & Autonomous Coding Agents, and Predictive Software Analytics. Her work combines machine learning, multi-objective evolutionary computation, and empirical software engineering to automatically make modern software systems fairer, greener, more reliable, and easier to maintain at industrial scale.

Verified Academic Profiles & Repositories
Google Scholar DBLP Bibliography UCL Institutional Profile GitHub (SOLAR-group) SOLAR Group Portal ACM Distinguished Speaker
UCL Engineering

Prof. Federica Sarro

Department of Computer Science

PhD • UCL SSE & SOLAR

Institutional Address
University College London
66-72 Gower Street, Bloomsbury
London WC1E 6EA, United Kingdom
Electronic Mail
f.sarro@ucl.ac.uk
ORCID Identifier
0000-0002-9146-442X
Society Membership
ACM Senior Member (since Oct. 2026)
IEEE Senior Member (since June 2026)
IEEE CS TCSE Chair (since Jan. 2025)
150+
Peer-Reviewed Papers
ICSE, FSE, ASE, TSE, TOSEM
Chair
IEEE CS TCSE
Executive Committee Leadership
2×
ACM Distinguished Paper
ICSE 2024 & ESEC/FSE
5×
Best Paper & HUMIES
ACM SIGEVO HUMIES Award
4×
Flagship PC Co-Chair
ASE '23, ICSME '21, ESEM '20, ICPC '19
2019
Founded SOLAR Group
Academic Biography & Career Trajectory

Scholarly Profile & Leadership

Federica Sarro is Full Professor of Software Engineering in the Department of Computer Science at University College London (UCL), where she serves as Head of the Software Systems Engineering (SSE) Group and Founder and Director of the SOLAR (Software Optimisation, Learning and Analytics Research) Group. She is also a senior member of the Centre for Research on Evolution, Search and Testing (CREST) and currently serves as elected Chair of the IEEE Computer Society Technical Community on Software Engineering (TCSE).

Academic Foundations & Rise at UCL

Professor Sarro completed her higher education in Computer Science at the Università degli Studi di Salerno, Italy, earning her Laurea Triennale (BSc), Laurea Specialistica (MSc, cum laude), and PhD in Computer Science. Following her doctoral research on empirical and search-based approaches to software project management and effort estimation, she joined the Department of Computer Science at University College London. Advancing rapidly through academic ranks due to her foundational contributions to automated software engineering, she was appointed Full Professor of Software Engineering in 2020.

Establishing SOLAR, UCLappA & Industrial Spin-Out Impact

In 2019, Professor Sarro established the SOLAR Research Group at UCL, bringing together post-doctoral fellows, PhD researchers, and industry collaborators around a unified mission: discovering automated ways to make software systems fairer, greener, more reliable, and easier to manage. Prior to and alongside SOLAR, she led the UCLappA (UCL App Store Analysis) group, whose pioneering empirical and predictive research on mobile software ecosystems directly led to the creation of the commercial spin-out company Appredict Ltd. Her research regularly transitions into real-world practice through collaborations with Bloomberg, Meta, Cisco, Microsoft, and Google.

International Scientific Service & Recognition

Professor Sarro has published over 150 peer-reviewed articles in premier software engineering venues including ICSE, ESEC/FSE, ASE, ISSTA, IEEE TSE, ACM TOSEM, and EMSE. Her work has been honoured with the IEEE TCSE Rising Star Award (2021), the Lero Parnas Fellowship (2023), two ACM Distinguished Paper Awards (including ICSE 2024), five Best Paper Awards, and an ACM SIGEVO HUMIES Award for human-competitive results in genetic and evolutionary computation. She has served as Program Co-Chair for flagship conferences including IEEE/ACM ASE 2023, IEEE ICSME 2021, IEEE/ACM ESEM 2020, and IEEE ICPC 2019.

Chronology

Academic Milestones

Salerno → UCL
  1. 2025 – 2026 Current

    Chair of IEEE CS TCSE & Head of UCL SSE

    Leading the global IEEE Technical Community on Software Engineering; landmark 2026 papers at ICSE & FSE on AI Multi-Agent Systems, ContextBench, and Bloomberg Bayesian analytics.

  2. 2023 – 2024

    Lero Parnas Fellow, ASE '23 PC Chair & ICSE '24 Award

    Awarded the prestigious Lero Parnas Fellowship; served as Program Co-Chair of IEEE/ACM ASE 2023; received the ACM Distinguished Paper Award at ICSE 2024 for intersectional software fairness.

  3. 2021 – 2022

    IEEE TCSE Rising Star Award & ICSME '21 PC Co-Chair

    Recognized with the IEEE TCSE Rising Star Award for foundational contributions to SBSE and software analytics; released Fairea (FSE '21) and MAAT (FSE '22).

  4. 2019 – 2020

    Full Professor at UCL & Founding of SOLAR Group

    Promoted to Full Professor of Software Engineering at UCL (2020); founded the SOLAR Research Group (2019); served as Program Co-Chair of ESEM 2020 and ICPC 2019.

  5. Prior – 2018

    Università degli Studi di Salerno, CREST & UCLappA

    Completed PhD, MSc and BSc at the University of Salerno; joined UCL CREST centre; led UCLappA and co-founded spin-out Appredict Ltd.

Research Portfolio & Recent Dispatches

Featured Research & 2025–2026 Highlights

Responsible AI & Fairness ICSE '24 Award

Algorithmic Fairness Testing & Automated Bias Repair

Pioneering mutation-based fairness benchmarking (Fairea), ensemble fairness-performance optimization (MAAT), intersectional protected attributes (FairMask), and search-based automatic bias repair.

LLMs & Coding Agents ICSE & FSE 2026

Multi-Agent Code Verification, ContextBench & Sustainable LLMs

Investigating industrial multi-agent systems for requirements-driven code verification (FSE 2026), context retrieval benchmarks for coding agents (ContextBench), LLM-guided Genetic Improvement, and empirical sustainability at NIER@ICSE 2026.

SBSE & Predictive Analytics Bloomberg & ERC

Multi-Objective Effort, Defect & Developer Experience Analytics

From CoGEE multi-objective effort estimation and MEG defect prediction ensembles to industrial deployment of BayesInsights and Pomona at Bloomberg (FSE 2026).

Recent News, Keynotes & Breakthroughs (2024–2026)

SOLAR & SSE Dispatch Log
2026
Multiple Papers Accepted at ICSE 2026 & FSE 2026

New breakthroughs including Empirical and Sustainability Aspects of SE Research in the Era of LLMs (NIER@ICSE '26), Industrial Deployment of an AI Multi-Agent System (FSE '26), and BayesInsights with Bloomberg (FSE '26).

2026
ContextBench & Pomona Continuous Code Quality Released

SOLAR introduces ContextBench for evaluating context retrieval in autonomous coding agents and collaborates with Bloomberg on Pomona for continuous automated code quality improvement.

2025
Keynote Lectures & LLM-Guided Genetic Improvement

Delivered international keynote addresses (including IEEE RE Keynote and ACM Distinguished Speaker lectures) and published Large Language Model Based Mutations in Genetic Improvement in the Automated Software Engineering journal.

2024
ACM Distinguished Paper Award at ICSE 2024 & TOSEM Survey

Awarded the ACM Distinguished Paper Award at ICSE 2024 for Fairness Improvement with Multiple Protected Attributes: How Far Are We? alongside the landmark Fairness Testing survey in ACM TOSEM.

Distinctions & Governance

Honours, Awards & Editorial Leadership

Major Honours & Awards

International Scientific Recognition

  • 2024
    ACM Distinguished Paper Award (ICSE 2024)
    46th IEEE/ACM International Conference on Software Engineering
  • 2023
    Lero Parnas Fellowship
    Science Foundation Ireland Research Centre for Software (Lero)
  • 2021
    IEEE TCSE Rising Star Award
    IEEE Computer Society Technical Council on Software Engineering
  • Multi
    2× ACM Distinguished Paper & 5× Best Paper Awards
    Recognized across premier SE & evolutionary computation venues
  • Award
    ACM SIGEVO HUMIES Award
    Human-Competitive Results Produced by Genetic & Evolutionary Computation
  • Active
    ACM Distinguished Speaker
    Association for Computing Machinery Distinguished Speaker Program

Editorial Boards

Flagship Journals in Software Engineering

  • ACM TOSEM Associate Editor
    ACM Transactions on Software Engineering and Methodology
  • IEEE TSE Associate Editor
    IEEE Transactions on Software Engineering
  • Springer EMSE Associate Editor
    Empirical Software Engineering Journal
  • Springer GPEM Associate Editor
    Genetic Programming and Evolvable Machines

Conference & Community Leadership

Executive & Program Chair Service

  • Chair
    IEEE CS Technical Community on Software Engineering (TCSE)
    Elected Chair of the global IEEE Computer Society TCSE
  • ASE '23
    Program Co-Chair, IEEE/ACM ASE 2023
    38th IEEE/ACM International Conference on Automated Software Engineering
  • ICSME '21
    Program Co-Chair, IEEE ICSME 2021
    37th IEEE International Conference on Software Maintenance and Evolution
  • ESEM '20
    Program Co-Chair, ACM/IEEE ESEM 2020
    14th ACM/IEEE International Symposium on Empirical Software Engineering and Measurement
  • ICPC '19
    Program Co-Chair, IEEE/ACM ICPC 2019
    27th IEEE/ACM International Conference on Program Comprehension
Industry Collaboration & Consultancy

Rigorous Research. Real-World Engineering Impact.

Connect empirical software engineering, multi-objective optimisation, and AI-enabled developer tools with industrial software teams. Explore published collaborations with Bloomberg, Meta, Microsoft, Google, JPMorgan Chase, and Cisco, including the B-Assist automated repair workflow.

Visit SOLAR
SOLAR Research Group • UCL Software Systems Engineering

Automated Ways to Make Software Fairer, Greener & More Reliable

Founded by Professor Federica Sarro in 2019 at University College London, the SOLAR (Software Optimisation, Learning and Analytics Research) group investigates the synergy of Search-Based Software Engineering (SBSE), Machine Learning, Large Language Models, and Empirical Software Engineering to solve high-impact challenges in modern software systems.

Visit SOLAR Lab Website (solar.cs.ucl.ac.uk) SOLAR-group Open Source
Research Architecture
The SOLAR Paradigm
Est. 2019
Software Optimisation: Multi-objective evolutionary search, hyper-heuristics & Genetic Improvement.
Software Learning: Fair & green ML systems, LLMs, and multi-agent verification pipelines.
Software Analytics: Effort estimation, defect prediction, Bayesian delivery models & app store mining.
Core Research Themes

Five Interactive Research Pillars

Pillar 01 • Responsible AI ICSE '24 Distinguished Paper

Software Fairness Testing & Automated Bias Mitigation

Machine learning software increasingly powers high-stakes societal decisions, yet frequently exhibits algorithmic bias across protected attributes. SOLAR pioneers automated fairness testing, mutation-based benchmarking (Fairea), ensemble fairness-accuracy optimization (MAAT), latent attribute translation (FairST), and search-based automatic repair for intersectional fairness.

Fairea (ESEC/FSE '21) MAAT (ESEC/FSE '22) Intersectional Fairness (ICSE '24) FairST (IST '26)
Key Venues: ICSE, FSE, TOSEM, EMSE
Pillar 02 • AI for SE ICSE 2026 & FSE 2026

LLMs, Multi-Agent Verification & Coding Agents

As Large Language Models and autonomous coding agents transform software development, rigorous evaluation and verification become paramount. Our research investigates industrial multi-agent systems for requirements-driven code verification (FSE 2026), repository-level context retrieval benchmarks (ContextBench), CLI agent trajectory analysis (TerminalWorld), and empirical sustainability reflections (NIER@ICSE 2026).

Multi-Agent Verification (FSE '26) ContextBench ('26) LLM Sustainability (ICSE-NIER '26) PR Acceptance (MSR '26)
Key Venues: ICSE, FSE, MSR, ASE
Pillar 03 • Evolutionary SE HUMIES & Bloomberg B-Assist

Search-Based Software Engineering & Automated Program Repair

Formulating software engineering tasks as multi-objective search and optimization problems. Research spans automated program repair in industrial pull-request workflows (B-Assist with Bloomberg), LLM-based mutations in Genetic Improvement (Gin), hyper-heuristic mutation testing (Sentinel), and repairing real-world API misuses.

B-Assist (Bloomberg × UCL) LLM Genetic Improvement ('25) Sentinel (IEEE TSE '20) CoGEE (ICSE '16)
Key Venues: ICSE, TSE, ASE, GECCO
Pillar 04 • Software Analytics Bloomberg & ERC Projects

Effort Estimation, Defect Prediction & Delivery Analytics

Harnessing historical repository telemetry, Bayesian networks, and multi-objective machine learning to support engineering decision-making. Highlights include modelling developer experience and software delivery at Bloomberg (BayesInsights, FSE 2026), multi-objective defect prediction ensembles (MEG), agile story-point estimation, and human-AI effort estimation.

BayesInsights (FSE '26) MEG Defect Ensembles (ESEM '22) Evaluation Measures (ISSTA '22) Human+ML Estimation (TSE '20)
Key Venues: FSE, ISSTA, TSE, ESEM
Pillar 05 • Sustainable & Mobile SE Green AI & UCLappA Spin-Out

Green Software Engineering, Deep Learning Efficiency & Mobile App Ecosystems

Investigating the environmental and non-functional costs of modern software systems—from empirically measuring the energy and run-time costs of deep learning frameworks (Green AI at ICSE 2022) and optimizing Android mobile performance via multi-objective Genetic Improvement (ACM TOSEM 2024, IEEE TSE 2022) to mining mobile app stores for requirements and user rating prediction (UCLappA / Appredict Ltd.).

Green AI Frameworks (ICSE '22) Android Genetic Improvement (TOSEM '24) Mobile Performance Survey (IEEE TSE '22) App Feature Rating Prediction (RE '18)
Key Venues: ICSE, TSE, TOSEM, IEEE RE
Open Science & Reproducibility

Flagship Open-Source Frameworks & Benchmarks

View All Repositories on GitHub (SOLAR-group)
Fairea ESEC/FSE '21

Model behaviour mutation framework for benchmarking bias mitigation methods and evaluating fairness-accuracy trade-offs in ML software.

SOLAR-group
MAAT & FairST FSE '22 / IST '26

Ensemble learning and latent sensitive attribute translation pipelines for addressing fairness and performance bugs in ML systems.

SOLAR-group
ContextBench Benchmark '26

Benchmark suite and evaluation harness for measuring context retrieval recall, token efficiency, and trajectory overhead in coding agents.

SOLAR-group
MEG & CoGEE ESEM '22 / ICSE '16

Multi-objective evolutionary algorithms for software defect prediction ensemble generation (MEG) and effort estimation (CoGEE).

SOLAR-group
Green AI & Android GI ICSE '22 / TOSEM '24

Empirical energy profiling benchmarks for deep learning frameworks and multi-objective Genetic Improvement for mobile apps.

SOLAR-group
Multi-Attr Fairness & Repair ICSE '24 / EMSE '24

Replication package and search-based repair algorithms for improving fairness across multiple intersectional protected attributes.

SOLAR-group
Industry Collaboration · Consultancy · Knowledge Exchange

Rigorous research. Real-world engineering impact.

Professor Federica Sarro’s research connects empirical evidence, multi-objective optimisation, and AI-enabled software engineering with the practical challenges faced by industrial software teams. Industry collaboration turns complex engineering bottlenecks into carefully evaluated methods, deployed developer tools, and actionable insights.

Discuss a Collaboration Read the Bloomberg Case Study
Research That Meets Industry

Shared problems. Stronger evidence. Better tools.

Professor Sarro has collaborated with global technology and financial organisations including AWS, Barclays, Bloomberg, Google, NEC, Meta, and JPMorgan Chase, as documented in UCL’s published profile of her work. UCL’s Software Systems Engineering (SSE) group—which Professor Sarro heads—also lists JPMorgan Chase and Cisco among its industry contributors, alongside the UCLappA spin-out company Appredict Ltd.

Click an organisation below to inspect collaborative context:

These names reflect published collaboration, research awards, and group-level industry links at UCL. They should not be read as a claim that every organisation is a current consultancy client, or as an endorsement by any company.

Example in Practice · Bloomberg × UCL Deployed Workflow Case Study

B-Assist: automated fixes, placed into developers’ existing workflow.

In a landmark collaboration with Bloomberg, UCL researchers helped develop B-Assist, which surfaces automated code-repair and static-analysis fix suggestions directly inside engineers’ existing pull-request workflows. The project illustrates how academic program-repair research can be shaped around engineers’ real practices and rigorously evaluated in an industrial environment, alongside follow-on collaborations such as BayesInsights (FSE 2026) and Pomona (2026).

74% of patch suggestions in an industrial evaluation study of 11 engineers were accepted outright or with small changes.
~80% rated suggestions useful or very useful inside their pull-request workflow, as reported in UCL’s published case study.
Read the full UCL case study
Bloomberg

Bloomberg × UCL Computer Science Partnership

Automated Program Repair · Bayesian Delivery Analytics · Code Quality

Multi-year industrial research collaboration with Bloomberg Engineering in London focusing on developer workflow automation, automated program repair (B-Assist), continuous small-change code quality improvement (Pomona), and probabilistic modelling of software delivery and developer experience using Bayesian Networks (BayesInsights, FSE 2026).

B-Assist Pull-Request Repair BayesInsights (FSE 2026) Pomona Continuous Quality (2026)
Enquire About Similar Work
Potential Areas of Collaboration

Where research can make a difference.

Projects work best when a clear engineering challenge meets a research question that can be investigated rigorously and responsibly.

⌘
01 · AI & Developer Tools

AI-enabled software engineering

Assessing AI-based coding support, autonomous multi-agent code verification, context retrieval (ContextBench), and automated program repair; examining how AI tools fit into existing developer workflows and how their real productivity and quality benefits should be measured.

AI for software Developer experience Automated repair Coding agents
⌁
02 · Analytics & Decision Support

Software analytics and prediction

Using software telemetry and repository data to investigate code quality, delivery bottlenecks, maintenance risk, and engineering effort (BayesInsights, CoGEE, MEG), with careful attention to empirical validation, causal context, and how predictions support engineering leadership decisions.

Predictive modelling Software metrics Decision support Bayesian networks
✳
03 · Quality & Optimisation

Testing, optimisation and repair

Exploring automated testing, Search-Based Software Engineering (SBSE), Genetic Improvement, test prioritisation, and program repair to navigate complex engineering design spaces and balance competing non-functional objectives (latency, energy, memory, reliability).

Testing Search-based methods Quality engineering Green AI efficiency
◎
04 · Evaluation & Trust

Evidence, fairness and responsible AI

Designing rigorous empirical evaluations and automated bias mitigation pipelines (Fairea, MAAT) that help teams understand model performance, intersectional fairness trade-offs, energy costs, and reliability—distinguishing promising demos from dependable production results.

Empirical studies Fairness & bias repair Trustworthiness Benchmarking
Ways to Work Together

Start with the challenge.

These are starting points for a conversation; scope, availability, and any formal arrangements are agreed through the appropriate UCL processes.

01 / APPLIED RESEARCH

Collaborative research projects

Define a shared research question, investigate it using relevant industrial data or workflows where appropriate, and agree on meaningful outputs for both engineering practice and peer-reviewed research.

Multi-year or focused sprints • UCL Research Services
02 / TECHNICAL REVIEW

Evaluation and consultancy discussions

Explore independent empirical assessment of a software tool, AI-enabled developer workflow, testing strategy, or predictive analytics approach—with a clear scope, evaluation criteria, and evidence base.

Advisory & technical evaluation • UCL Consultants (UCLC)
03 / TALENT & KNOWLEDGE

Industry-linked student projects

Discuss research-aligned MSc dissertations, industrially sponsored PhD studentships (e.g., UKRI CASE / industry fellowships), or visiting researcher placements investigating real engineering challenges.

PhD / MSc supervision & Academic Visitors
04 / KNOWLEDGE EXCHANGE

Research briefings and workshops

Explore executive briefings, invited keynotes, and technical workshops around software analytics, optimisation, program repair, empirical evaluation, and responsible AI for engineering leaders.

Executive seminars & R&D workshops
Prospective Team Members & Talent Pipeline

Joining the SOLAR Research Group (MSc, PhD & Visitors)

SOLAR: Work With Us Guidelines
01 · MSc Students

Explore a Research Project

Considering a research-led MSc dissertation at UCL? Relevant themes include software analytics, predictive modelling, automated testing, and search-based optimisation. Include your CV, academic transcript, research interests, and preferred start period.

Enquire about an MSc project
02 · PhD Candidates

Develop a Research Question

Prospective PhD researchers are encouraged to describe the questions they want to explore in SBSE, predictive analytics, testing, repair, coding agents, or software fairness. Include your CV, research statement, transcripts, and ideal start time.

Discuss PhD research
03 · Academic & Industry Visitors

Explore a Research Visit

SOLAR welcomes expressions of interest from visiting scholars and industrial fellows. Proposals are strongest when they identify a shared research question, collaboration plan, preferred visit dates/duration, and funding arrangements.

Enquire about a visit
Start an Enquiry

Have a real engineering problem to explore?

Send a concise outline of the engineering challenge, the team or software system involved, and what a useful outcome would look like. An initial conversation helps determine whether there is a strong research or consultancy fit.

Direct Contact Coordinates
Electronic Mail: f.sarro@ucl.ac.uk
Department: UCL Computer Science (Gower St)

This page describes potential collaboration themes, not a guaranteed consultancy service or confirmation of current project capacity. Any engagement is subject to mutual agreement and applicable UCL policies.

Structured Collaboration & Consultancy Brief f.sarro@ucl.ac.uk
Prospective Doctoral Researchers • UELA Scholarships

Open PhD Positions

Collaborative Doctoral Training

Three UELA funded collaborative project PhD scholarships (start date of 1 October 2027)

Deadline: 16/12/2026
  • Barclays PIs: Ye, Sarro

    "Exploring novel ways to drive customer and merchant benefits through huge-scale transaction data"

    See project details
  • Bloomberg PIs: Petke, Sarro

    "Specification- and Verification-First Human-in-the-Loop Software Factories with Agentic AI"

    See project details
  • NEC Laboratories Europe PI: Sarro

    "Addressing the Cognitive Gap in Software Development: Developer Tools for Keeping Human Understanding Up to Speed with AI Coding Agents"

    See project details

See more info on funds, requirements and how to apply here.

Application deadline: 16/12/2026.

UELA Studentships Portal
Scholarly Bibliography & Live Archive

Publications

Below is an expanded interactive library of key peer-reviewed contributions. For my complete list of publications and citation metrics, access her official live profiles:

Google Scholar (Full Record of 150+ Works) ↗ DBLP Computer Science Bibliography ↗ ORCID Record ↗
Showing publications