Sr Software Engineer - Coding Agents

Lexsi Labs; Mumbai; In loco; INR 2.800.000 - 4.600.000; Tempo pieno

Descrizione del lavoro

Lexsi Labs in Mumbai, India, is building aligned, interpretable, and safe autonomous systems. This developer tooling role focuses on the agent architecture, tools surface, and integrations that engineers rely on every day.

You will work with a research team on evaluation design, post-training interpretability, and production-grade tooling for complex, evergreen codebases, with strong ownership and fast execution.

Competenze

• This is a developer tooling role requiring hands-on experience with tooling and IDE internals.
• Experience with code intelligence, ASTs, and static analysis is preferred.
• Familiarity with build systems and CI/CD pipelines is expected.
• Knowledge of observability, tracing, and debugging of non-deterministic systems is a plus.

Mansioni

• Own the agent architecture surface and tooling for software work (retrieval, editing, builds, tests, static analysis, VCS).
• Implement verification that changes are correct before humans review them.
• Support long-horizon experiments across migrations with checkpointing and recovery.
• Develop the authorization and rollback model governing agent actions and undo behavior.
• Integrate with engineering stacks and observability tooling used to debug and audit agents.
• Build evaluation suites from repository tasks and tune agent behavior accordingly.

Conoscenze

Developer tooling

IDEs

Code intelligence

Build systems

CI/CD

Observability

Integrations

Performance tuning

Safety and alignment

Interpretability

LangChain/agents awareness

Strumenti

Jest/Testing frameworks

Build tools

Static analysis tools

Descrizione del lavoro

LexsiLabsistheleadingfrontierAIlabfocusedonbuilding aligned,interpretable,andsafesuperintelligentsystems .Whilethatisthevision,themissiontobuildsafetyawareautonomoussystemintheextremenearterm.OurresearchworkspansaroundareaslikeAIalignmentmethodologies,interpretability ledsystemdesign,andfoundationalmodelresearchacrossstructured,tabular,andnewautonomoussystemdesigns.Wepublishedabout25+papersinthepast15monthsacrossleadingconferencesICLR,ICML,WWW,IJCNN,MICCAI,Euripsetc.OurlabsarelocatedinIndia(Mumbai&remote),Paris,&London.

Weoperatewithaflatstructure,highautonomy,andastrongbiastowardengineerswhotakefullownershipofwhattheybuild,fromarchitecturetoproductionbehavior.

TheRole

Ourcurrentsprintonbuildingautonomoussystemforcomplexproblems,acrosssoftwareengineering,datascience,andAIresearchinvolvesbuildingharness,executionsubstrateandevaluationsystem,andeachisdesignedtoruninsideacustomer'senvironmentratherthanours.

Thisroleownsthecodingagent:asystemthattakesanobjective,worksdirectlyonarepository,verifiesitsownchanges,andproducesarecordofwhatitdid.Thecodebasesarelarge,old,thinlytestedandload-bearing,andtheagentrunsinsidethecustomer'snetwork,usuallywithoutinternetegress.

What'llyouworkon:

• Theagentarchitectureanditstoolsurfaceforsoftwarework,includingretrieval,editing,builds,tests,staticanalysisandversioncontrol.
• Verification,sothatachangeisestablishedascorrectbeforeahumanseesit.
• Long-horizonexecutionacrossmulti-filerefactorsandmigrations,withcheckpointing,recoveryandself-correction.
• Theauthorisationandrollbackmodelgoverningwhattheagentmaydoandhowitisundone.
• Integrationswiththeengineeringestatetheagentoperatesin,andtheobservabilitytoolingusedtodebugandauditit.
• Evaluationsuitesbuiltfromrealrepositorytasks,andthebehaviortuningthatfollowsfromthem.

Youwillworkcloselywithourresearchteamonevaluationdesign,post-trainingandinterpretabilityofagentbehavior.

WhatWeAreLookingFor

ThisisadevelopertoolingroleasmuchasanAIrole.Theagentisonlyascapableasthetoolsandintegrationsitisgiven,sodepthheremattersmoretousthananythingelse.

• DeepexperiencewithdevelopertoolingandIDEs.EditorandIDEinternalsorextensions,languageserversandLSP,codeintelligence,refactoringengines.Youhavebuiltthingsengineersuseeveryday.
• Codeintelligenceandprogramanalysis.ASTsandtree-sitter,staticanddataflowanalysis,symbolindexingandcodesearch,callgraphanddependencyresolution,codemodsandautomatedmigrationtooling.
• BuildsystemsandCI/CD.Bazel,Gradle,Maven,Makeandsimilar,incrementalandhermeticbuilds,artifactanddependencymanagement,pipelinedesignacrossJenkins,GitLab,GitHubActionsorequivalents,includingself-hostedandofflinedeployments.
• Observability.Distributedtracing,structuredlogging,OpenTelemetry,andbuildingthereplayandinspectiontoolingthatmakesnon-deterministicsystemsdebuggable.
• Integrationsengineering.GithostingAPIs,ticketingandworkflowsystems,enterpriseauth,andtherealityofworkingagainstself-hostedsoftwareseveralversionsbehindcurrent.
• Agenticsystemsinproduction,notprototypes.Youknowthecurrentframeworklandscape(ReAct-styleagents,LangGraph,LangChain,SemanticKernel)andwhereeachonestopsworking.
• Youreasonaboutperformance,reliability,cost,safetyandinterpretabilityasoneconnectedsetofconstraints.
• Youmakegoodcallsinthegrayarea,andyouarecomfortablewhenproblemsarelooselyspecifiedandownershipisassumedratherthanassigned.
• ExperiencebuildingevaluationdatasetsorframeworksforAIsystemsisaplus.
• Compilerbackground,largemonorepowork,orlegacymodernisationprogramsareaplus.
• Priorexposuretoalignment,interpretabilityorsafetytoolinginproductionsettingsisaplus.

Wemovequicklyandexpectcandidatestodothesame.Wevaluesubstanceoverpolishandexecutionoverrhetoric.