Job application tracker and AI-powered job search assistant. Helps job seekers manage their search journey with AI resume review, job matching, task logging, and application analytics, all while keeping your data private.
TypeScript ai-agentai-resume-analyzerapplication-trackerapplication-tracking-systematscareerjob-applicationjob-automationjob-huntingjob-searchjob-seeker-toolsjob-trackerjob-trackingjobsearchjobseekersmcpollamaresumeresume-reviewself-hosted
add resume preprocessing step
Khuram Niaza6a58c3f4edf
diff --git a/src/app/api/ai/resume/review/route.ts b/src/app/api/ai/resume/review/route.tsindex abb742e..7c426e7 100644--- a/src/app/api/ai/resume/review/route.ts+++ b/src/app/api/ai/resume/review/route.ts@@ -14,8 +14,8 @@ import {getKeywordCountFromSemantic,getVerbCountFromSemantic,AIUnavailableError,+ preprocessResume,} from "@/lib/ai";-import { convertResumeToText } from "@/utils/ai.utils";import { Resume } from "@/models/profile.model";import { AiModel } from "@/models/ai.model";@@ -56,19 +56,24 @@ export const POST = async (req: NextRequest) => {}try {- const resumeText = await convertResumeToText(resume);-- // ...existing code...+ const preprocessResult = await preprocessResume(resume);+ if (!preprocessResult.success) {+ return NextResponse.json(+ { error: preprocessResult.error.message, code: preprocessResult.error.code },+ { status: 400 },+ );+ }+ const { normalizedText, metadata } = preprocessResult.data;// Use semantic extraction for keywords and verbsconst [semanticKeywords, semanticVerbs] = await Promise.all([extractSemanticKeywords(- resumeText,+ normalizedText,selectedModel.provider,selectedModel.model || "llama3.2",),analyzeActionVerbs(- resumeText,+ normalizedText,selectedModel.provider,selectedModel.model || "llama3.2",),@@ -82,7 +87,7 @@ export const POST = async (req: NextRequest) => {];const allVerbs = semanticVerbs.strong_verbs.map((v) => v.verb);- const basePrompt = buildResumeReviewPrompt(resumeText);+ const basePrompt = buildResumeReviewPrompt(normalizedText);const fullPrompt = `${basePrompt}TOOL ANALYSIS RESULTS (use these in your evaluation):diff --git a/src/lib/ai/index.ts b/src/lib/ai/index.tsindex 473638a..c71e43a 100644--- a/src/lib/ai/index.ts+++ b/src/lib/ai/index.ts@@ -18,6 +18,15 @@ export {// Analysis toolsexport { AIUnavailableError } from "./tools";+// Preprocessing+export {+ preprocessResume,+ convertResumeToText,+ type PreprocessingResult,+ type ResumeMetadata,+ type PreprocessedResume,+} from "./tools";+// Semantic extraction and similarity functionsexport {extractSemanticKeywords,diff --git a/src/lib/ai/tools/index.ts b/src/lib/ai/tools/index.tsindex 70ea7ab..d43e6d8 100644--- a/src/lib/ai/tools/index.ts+++ b/src/lib/ai/tools/index.ts@@ -31,3 +31,12 @@ export {calculateSemanticSimilarity,generateMatchExplanation,} from "./semantic-similarity";++// Preprocessing+export {+ preprocessResume,+ convertResumeToText,+ type PreprocessingResult,+ type ResumeMetadata,+ type PreprocessedResume,+} from "./preprocessing";diff --git a/src/lib/ai/tools/preprocessing.ts b/src/lib/ai/tools/preprocessing.tsnew file mode 100644index 0000000..37604b6--- /dev/null+++ b/src/lib/ai/tools/preprocessing.ts@@ -0,0 +1,322 @@+/**+ * Resume Preprocessing Module+ * Normalizes text, detects sections, extracts metadata, and validates resume content+ */++import {+ ContactInfo,+ Education,+ Resume,+ ResumeSection,+ SectionType,+ WorkExperience,+} from "@/models/profile.model";++// TYPES++export interface ResumeMetadata {+ characterCount: number;+ wordCount: number;+ lineCount: number;+ hasContactInfo: boolean;+}++export interface PreprocessedResume {+ normalizedText: string;+ metadata: ResumeMetadata;+ isValid: boolean;+}++export type PreprocessingResult =+ | { success: true; data: PreprocessedResume }+ | {+ success: false;+ error: { code: string; message: string; details?: object };+ };++// VALIDATION THRESHOLDS++const MIN_WORD_COUNT = 50;+const MIN_CHAR_COUNT = 200;+const MAX_WORD_COUNT = 10000;+const MAX_CONSECUTIVE_SPECIAL_CHARS = 20;++// UTILITY FUNCTIONS++export const removeHtmlTags = (description: string | undefined): string => {+ if (!description) return "";++ return description+ .replace(/<li[^>]*>/gi, "• ")+ .replace(/<\/(li|p|div|br)[^>]*>/gi, "\n")+ .replace(/<br\s*\/?>/gi, "\n")+ .replace(/<[^>]+>/g, " ")+ .replace(/\s+/g, " ")+ .replace(/\n\s*\n/g, "\n")+ .trim();+};++// NORMALIZATION FUNCTIONS++export const normalizeWhitespace = (text: string): string => {+ return text+ .replace(/\r\n/g, "\n")+ .replace(/\r/g, "\n")+ .replace(/[ \t]+/g, " ")+ .replace(/\n{3,}/g, "\n\n")+ .trim();+};++export const normalizeBullets = (text: string): string => {+ return text+ .replace(/[•●○◦▪▸►◆★✦✓✔→‣⁃]/g, "•")+ .replace(/^[-–—]\s/gm, "• ")+ .replace(/^\*\s/gm, "• ");+};++export const normalizeHeadings = (text: string): string => {+ return text+ .replace(/^([A-Z][A-Z\s&]+):?\s*$/gm, (_match, heading) => {+ const normalized = heading.trim().replace(/:$/, "");+ return `\n${normalized}\n`;+ })+ .replace(/\n{3,}/g, "\n\n");+};++// METADATA EXTRACTION++export const extractMetadata = (text: string): ResumeMetadata => {+ const words = text.split(/\s+/).filter((w) => w.length > 0);+ const lines = text.split("\n");++ return {+ characterCount: text.length,+ wordCount: words.length,+ lineCount: lines.length,+ hasContactInfo: hasContactPatterns(text),+ };+};++const hasContactPatterns = (text: string): boolean => {+ const emailPattern = /[\w.-]+@[\w.-]+\.\w+/;+ const phonePattern = /\(?\d{3}\)?[-.\s]?\d{3}[-.\s]?\d{4}/;+ return emailPattern.test(text) || phonePattern.test(text);+};++// VALIDATION++interface ValidationResult {+ isValid: boolean;+ error?: { code: string; message: string; details?: object };+}++export const validateResume = (+ text: string,+ metadata: ResumeMetadata,+): ValidationResult => {+ // Check for empty content+ if (!text || text.trim().length === 0) {+ return {+ isValid: false,+ error: {+ code: "NO_CONTENT",+ message: "Resume appears to be empty or contains only whitespace",+ },+ };+ }++ // Check minimum length+ if (+ metadata.wordCount < MIN_WORD_COUNT ||+ metadata.characterCount < MIN_CHAR_COUNT+ ) {+ return {+ isValid: false,+ error: {+ code: "TOO_SHORT",+ message: `Resume is too short. Found ${metadata.wordCount} words and ${metadata.characterCount} characters. Minimum required: ${MIN_WORD_COUNT} words or ${MIN_CHAR_COUNT} characters.`,+ details: {+ wordCount: metadata.wordCount,+ characterCount: metadata.characterCount,+ minWordCount: MIN_WORD_COUNT,+ minCharCount: MIN_CHAR_COUNT,+ },+ },+ };+ }++ // Check for corruption - too long+ if (metadata.wordCount > MAX_WORD_COUNT) {+ return {+ isValid: false,+ error: {+ code: "CORRUPTED",+ message: `Resume appears to be corrupted or contains excessive content. Found ${metadata.wordCount} words, maximum allowed: ${MAX_WORD_COUNT}.`,+ details: {+ wordCount: metadata.wordCount,+ maxWordCount: MAX_WORD_COUNT,+ },+ },+ };+ }++ // Check for corruption - consecutive special characters+ const specialCharPattern = new RegExp(+ `[^a-zA-Z0-9\\s]{${MAX_CONSECUTIVE_SPECIAL_CHARS + 1},}`,+ );+ if (specialCharPattern.test(text)) {+ return {+ isValid: false,+ error: {+ code: "CORRUPTED",+ message:+ "Resume appears to be corrupted. Found excessive consecutive special characters.",+ details: {+ maxConsecutiveSpecialChars: MAX_CONSECUTIVE_SPECIAL_CHARS,+ },+ },+ };+ }++ return { isValid: true };+};++// RESUME TO TEXT CONVERSION (moved from ai.utils.ts)++export const convertResumeToText = (resume: Resume): Promise<string> => {+ return new Promise((resolve) => {+ const formatContactInfo = (contactInfo?: ContactInfo) => {+ if (!contactInfo) return "";+ const parts = [+ `Name: ${contactInfo.firstName} ${contactInfo.lastName}`,+ contactInfo.headline ? `Headline: ${contactInfo.headline}` : "",+ contactInfo.email ? `Email: ${contactInfo.email}` : "",+ contactInfo.phone ? `Phone: ${contactInfo.phone}` : "",+ contactInfo.address ? `Address: ${contactInfo.address}` : "",+ ].filter(Boolean);+ return parts.join("\n");+ };++ const formatWorkExperiences = (workExperiences?: WorkExperience[]) => {+ if (!workExperiences || workExperiences.length === 0) return "";+ return workExperiences+ .map((experience) => {+ const desc = removeHtmlTags(experience.description);+ const parts = [+ `Company: ${experience.Company.label}`,+ `Job Title: ${experience.jobTitle.label}`,+ `Location: ${experience.location.label}`,+ desc ? `Description: ${desc}` : "",+ ].filter(Boolean);+ return parts.join("\n");+ })+ .join("\n\n");+ };++ const formatEducation = (educations?: Education[]) => {+ if (!educations || educations.length === 0) return "";+ return educations+ .map((education) => {+ const desc = removeHtmlTags(education.description);+ const parts = [+ `Institution: ${education.institution}`,+ `Degree: ${education.degree}`,+ `Field of Study: ${education.fieldOfStudy}`,+ `Location: ${education.location.label}`,+ desc ? `Description: ${desc}` : "",+ ].filter(Boolean);+ return parts.join("\n");+ })+ .join("\n\n");+ };++ const formatResumeSections = (sections?: ResumeSection[]) => {+ if (!sections || sections.length === 0) return "";+ return sections+ .map((section) => {+ switch (section.sectionType) {+ case SectionType.SUMMARY:+ return `Summary: ${removeHtmlTags(section.summary?.content)}`;+ case SectionType.EXPERIENCE:+ return formatWorkExperiences(section.workExperiences);+ case SectionType.EDUCATION:+ return formatEducation(section.educations);+ default:+ return "";+ }+ })+ .join("\n");+ };++ const inputMessage = `+Title: ${resume.title}+${formatContactInfo(resume.ContactInfo)}+${formatResumeSections(resume.ResumeSections)}+`;+ return resolve(inputMessage);+ });+};++// MAIN ORCHESTRATOR++export const preprocessResume = async (+ resume: Resume,+): Promise<PreprocessingResult> => {+ try {+ // Convert resume object to raw text+ const rawText = await convertResumeToText(resume);++ // Quick validation - fail fast if obviously invalid+ if (!rawText || rawText.trim().length < MIN_CHAR_COUNT) {+ const charCount = rawText?.trim().length || 0;+ return {+ success: false,+ error: {+ code: charCount === 0 ? "NO_CONTENT" : "TOO_SHORT",+ message:+ charCount === 0+ ? "Resume appears to be empty"+ : `Resume is too short (${charCount} characters, minimum ${MIN_CHAR_COUNT} required)`,+ details: { characterCount: charCount },+ },+ };+ }++ // Apply normalization pipeline+ let normalizedText = rawText;+ normalizedText = normalizeWhitespace(normalizedText);+ normalizedText = normalizeBullets(normalizedText);+ normalizedText = normalizeHeadings(normalizedText);++ // Extract metadata+ const metadata = extractMetadata(normalizedText);++ // Full validation+ const validationResult = validateResume(normalizedText, metadata);+ console.log("RESUME: ", { normalizedText, metadata });+ if (!validationResult.isValid) {+ return {+ success: false,+ error: validationResult.error!,+ };+ }++ return {+ success: true,+ data: {+ normalizedText,+ metadata,+ isValid: true,+ },+ };+ } catch (error) {+ const message = error instanceof Error ? error.message : "Unknown error";+ return {+ success: false,+ error: {+ code: "PREPROCESSING_ERROR",+ message: `Failed to preprocess resume: ${message}`,+ },+ };+ }+};diff --git a/src/utils/ai.utils.ts b/src/utils/ai.utils.tsindex 40c1bb4..89b3964 100644--- a/src/utils/ai.utils.ts+++ b/src/utils/ai.utils.ts@@ -1,27 +1,18 @@import { JobResponse } from "@/models/job.model";-import {- ContactInfo,- Education,- Resume,- ResumeSection,- SectionType,- WorkExperience,-} from "@/models/profile.model";import { AiProvider } from "@/models/ai.model";+// Re-export for backwards compatibility+export { convertResumeToText } from "@/lib/ai/tools/preprocessing";+const removeHtmlTags = (description: string | undefined): string => {if (!description) return "";return (description- // Convert list items to bullet points before stripping tags.replace(/<li[^>]*>/gi, "• ")- // Add line breaks after block elements.replace(/<\/(li|p|div|br)[^>]*>/gi, "\n").replace(/<br\s*\/?>/gi, "\n")- // Remove all remaining HTML tags.replace(/<[^>]+>/g, " ")- // Clean up excessive whitespace while preserving structure.replace(/\s+/g, " ").replace(/\n\s*\n/g, "\n").trim()@@ -132,87 +123,6 @@ export const fetchRunningModels = async (): Promise<{}};-export const convertResumeToText = (resume: Resume): Promise<string> => {- return new Promise((resolve) => {- const formatContactInfo = (contactInfo?: ContactInfo) => {- if (!contactInfo) return "";- const parts = [- `Name: ${contactInfo.firstName} ${contactInfo.lastName}`,- contactInfo.headline ? `Headline: ${contactInfo.headline}` : "",- contactInfo.email ? `Email: ${contactInfo.email}` : "",- contactInfo.phone ? `Phone: ${contactInfo.phone}` : "",- contactInfo.address ? `Address: ${contactInfo.address}` : "",- ].filter(Boolean);- return parts.join("\n");- };-- const formatWorkExperiences = (workExperiences?: WorkExperience[]) => {- if (!workExperiences || workExperiences.length === 0) return "";- return workExperiences- .map((experience) => {- const desc = removeHtmlTags(experience.description);- const parts = [- `Company: ${experience.Company.label}`,- `Job Title: ${experience.jobTitle.label}`,- `Location: ${experience.location.label}`,- desc ? `Description: ${desc}` : "",- ].filter(Boolean);- return parts.join("\n");- })- .join("\n\n");- };- // Start Date: ${experience.startDate.toLocaleDateString().split("T")[0]}- // End Date: ${- // experience.currentJob- // ? "Present"- // : experience.endDate.toLocaleDateString().split("T")[0]- // }-- const formatEducation = (educations?: Education[]) => {- if (!educations || educations.length === 0) return "";- return educations- .map((education) => {- const desc = removeHtmlTags(education.description);- const parts = [- `Institution: ${education.institution}`,- `Degree: ${education.degree}`,- `Field of Study: ${education.fieldOfStudy}`,- `Location: ${education.location.label}`,- desc ? `Description: ${desc}` : "",- ].filter(Boolean);- return parts.join("\n");- })- .join("\n\n");- };-- const formatResumeSections = (sections?: ResumeSection[]) => {- if (!sections || sections.length === 0) return "";- return sections- .map((section) => {- switch (section.sectionType) {- case SectionType.SUMMARY:- return `Summary: ${removeHtmlTags(section.summary?.content)}`;- case SectionType.EXPERIENCE:- return formatWorkExperiences(section.workExperiences);- case SectionType.EDUCATION:- return formatEducation(section.educations);- default:- return "";- }- })- .join("\n");- };-- const inputMessage = `-- Title: ${resume.title}- ${formatContactInfo(resume.ContactInfo)}- ${formatResumeSections(resume.ResumeSections)}- `;- return resolve(inputMessage);- });-};-export const convertJobToText = (job: JobResponse): Promise<string> => {return new Promise((resolve) => {const {